🍻 cheers
Browse files- README.md +5 -4
- all_results.json +10 -10
- eval_results.json +6 -6
- runs/Mar26_20-28-31_ryanserver/events.out.tfevents.1711503391.ryanserver.7179.1 +3 -0
- train_results.json +4 -4
- trainer_state.json +1600 -595
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
model-index:
|
| 7 |
- name: ryan_model314
|
|
@@ -13,11 +14,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# ryan_model314
|
| 15 |
|
| 16 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss: 0.
|
| 19 |
-
- Na Accuracy: 0.
|
| 20 |
-
- Ordinal Accuracy: 0.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
model-index:
|
| 8 |
- name: ryan_model314
|
|
|
|
| 14 |
|
| 15 |
# ryan_model314
|
| 16 |
|
| 17 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the beans dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.2532
|
| 20 |
+
- Na Accuracy: 0.947
|
| 21 |
+
- Ordinal Accuracy: 0.5952
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
"eval_accuracy": 0.628,
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_na_accuracy": 0.
|
| 6 |
-
"eval_ordinal_accuracy": 0.
|
| 7 |
-
"eval_runtime":
|
| 8 |
-
"eval_samples_per_second":
|
| 9 |
-
"eval_steps_per_second": 3.
|
| 10 |
-
"train_loss": 0.
|
| 11 |
-
"train_runtime":
|
| 12 |
-
"train_samples_per_second":
|
| 13 |
-
"train_steps_per_second": 0.
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
"eval_accuracy": 0.628,
|
| 4 |
+
"eval_loss": 0.2531912922859192,
|
| 5 |
+
"eval_na_accuracy": 0.947,
|
| 6 |
+
"eval_ordinal_accuracy": 0.5951557093425606,
|
| 7 |
+
"eval_runtime": 41.4892,
|
| 8 |
+
"eval_samples_per_second": 24.103,
|
| 9 |
+
"eval_steps_per_second": 3.013,
|
| 10 |
+
"train_loss": 0.15650403581261635,
|
| 11 |
+
"train_runtime": 3981.9696,
|
| 12 |
+
"train_samples_per_second": 10.045,
|
| 13 |
+
"train_steps_per_second": 0.628
|
| 14 |
}
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_na_accuracy": 0.
|
| 5 |
-
"eval_ordinal_accuracy": 0.
|
| 6 |
-
"eval_runtime":
|
| 7 |
-
"eval_samples_per_second":
|
| 8 |
-
"eval_steps_per_second": 3.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"eval_loss": 0.2531912922859192,
|
| 4 |
+
"eval_na_accuracy": 0.947,
|
| 5 |
+
"eval_ordinal_accuracy": 0.5951557093425606,
|
| 6 |
+
"eval_runtime": 41.4892,
|
| 7 |
+
"eval_samples_per_second": 24.103,
|
| 8 |
+
"eval_steps_per_second": 3.013
|
| 9 |
}
|
runs/Mar26_20-28-31_ryanserver/events.out.tfevents.1711503391.ryanserver.7179.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79bb34c6fad13e7161dccdd423749c0e0c7f10a793282d952ea4b3f94316cd47
|
| 3 |
+
size 474
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.15650403581261635,
|
| 4 |
+
"train_runtime": 3981.9696,
|
| 5 |
+
"train_samples_per_second": 10.045,
|
| 6 |
+
"train_steps_per_second": 0.628
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1024 +1,2029 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./ryan_model314/checkpoint-
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 0.
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 0.
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 0.
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 0.
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 1.
|
| 42 |
-
"learning_rate": 0.
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"grad_norm": 1.
|
| 56 |
-
"learning_rate": 0.
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
-
"epoch": 0.
|
| 62 |
-
"grad_norm": 1.
|
| 63 |
-
"learning_rate": 0.
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
-
"epoch": 0.
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate": 0.
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 0.
|
| 76 |
-
"grad_norm": 1.
|
| 77 |
-
"learning_rate": 0.
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
-
"epoch": 0.
|
| 83 |
-
"eval_loss": 0.
|
| 84 |
-
"eval_na_accuracy": 0.
|
| 85 |
-
"eval_ordinal_accuracy": 0.
|
| 86 |
-
"eval_runtime":
|
| 87 |
-
"eval_samples_per_second": 9.
|
| 88 |
-
"eval_steps_per_second": 1.
|
| 89 |
"step": 100
|
| 90 |
},
|
| 91 |
{
|
| 92 |
-
"epoch": 0.
|
| 93 |
-
"grad_norm": 1.
|
| 94 |
-
"learning_rate": 0.
|
| 95 |
-
"loss": 0.
|
| 96 |
"step": 110
|
| 97 |
},
|
| 98 |
{
|
| 99 |
-
"epoch": 0.
|
| 100 |
-
"grad_norm": 0.
|
| 101 |
-
"learning_rate": 0.
|
| 102 |
-
"loss": 0.
|
| 103 |
"step": 120
|
| 104 |
},
|
| 105 |
{
|
| 106 |
-
"epoch": 0.
|
| 107 |
-
"grad_norm":
|
| 108 |
-
"learning_rate": 0.
|
| 109 |
-
"loss": 0.
|
| 110 |
"step": 130
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch": 0.
|
| 114 |
-
"grad_norm":
|
| 115 |
-
"learning_rate": 0.
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 140
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"grad_norm": 0.
|
| 122 |
-
"learning_rate": 0.
|
| 123 |
-
"loss": 0.
|
| 124 |
"step": 150
|
| 125 |
},
|
| 126 |
{
|
| 127 |
-
"epoch": 0.
|
| 128 |
-
"grad_norm": 1.
|
| 129 |
-
"learning_rate": 0.
|
| 130 |
-
"loss": 0.
|
| 131 |
"step": 160
|
| 132 |
},
|
| 133 |
{
|
| 134 |
-
"epoch": 0.
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate": 0.
|
| 137 |
-
"loss": 0.
|
| 138 |
"step": 170
|
| 139 |
},
|
| 140 |
{
|
| 141 |
-
"epoch": 0.
|
| 142 |
-
"grad_norm":
|
| 143 |
-
"learning_rate": 0.
|
| 144 |
-
"loss": 0.
|
| 145 |
"step": 180
|
| 146 |
},
|
| 147 |
{
|
| 148 |
-
"epoch": 0.
|
| 149 |
-
"grad_norm":
|
| 150 |
-
"learning_rate": 0.
|
| 151 |
-
"loss": 0.
|
| 152 |
"step": 190
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch": 0.
|
| 156 |
-
"grad_norm": 0.
|
| 157 |
-
"learning_rate": 0.
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 200
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"eval_loss": 0.
|
| 164 |
-
"eval_na_accuracy": 0.
|
| 165 |
-
"eval_ordinal_accuracy": 0.
|
| 166 |
-
"eval_runtime":
|
| 167 |
-
"eval_samples_per_second": 25.
|
| 168 |
-
"eval_steps_per_second": 3.
|
| 169 |
"step": 200
|
| 170 |
},
|
| 171 |
{
|
| 172 |
-
"epoch": 0.
|
| 173 |
-
"grad_norm": 1.
|
| 174 |
-
"learning_rate": 0.
|
| 175 |
-
"loss": 0.
|
| 176 |
"step": 210
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch": 0.
|
| 180 |
-
"grad_norm": 1.
|
| 181 |
-
"learning_rate": 0.
|
| 182 |
-
"loss": 0.
|
| 183 |
"step": 220
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"grad_norm":
|
| 188 |
-
"learning_rate": 0.
|
| 189 |
-
"loss": 0.
|
| 190 |
"step": 230
|
| 191 |
},
|
| 192 |
{
|
| 193 |
-
"epoch": 0.
|
| 194 |
-
"grad_norm":
|
| 195 |
-
"learning_rate": 0.
|
| 196 |
-
"loss": 0.
|
| 197 |
"step": 240
|
| 198 |
},
|
| 199 |
{
|
| 200 |
-
"epoch": 0.
|
| 201 |
-
"grad_norm":
|
| 202 |
-
"learning_rate": 0.
|
| 203 |
-
"loss": 0.
|
| 204 |
"step": 250
|
| 205 |
},
|
| 206 |
{
|
| 207 |
-
"epoch": 0.
|
| 208 |
-
"grad_norm":
|
| 209 |
-
"learning_rate": 0.
|
| 210 |
-
"loss": 0.
|
| 211 |
"step": 260
|
| 212 |
},
|
| 213 |
{
|
| 214 |
-
"epoch": 0.
|
| 215 |
-
"grad_norm": 0.
|
| 216 |
-
"learning_rate": 0.
|
| 217 |
-
"loss": 0.
|
| 218 |
"step": 270
|
| 219 |
},
|
| 220 |
{
|
| 221 |
-
"epoch": 0.
|
| 222 |
-
"grad_norm":
|
| 223 |
-
"learning_rate": 0.
|
| 224 |
-
"loss": 0.
|
| 225 |
"step": 280
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch": 0.
|
| 229 |
-
"grad_norm":
|
| 230 |
-
"learning_rate": 0.
|
| 231 |
-
"loss": 0.
|
| 232 |
"step": 290
|
| 233 |
},
|
| 234 |
{
|
| 235 |
-
"epoch": 0.
|
| 236 |
-
"grad_norm":
|
| 237 |
-
"learning_rate": 0.
|
| 238 |
-
"loss": 0.
|
| 239 |
"step": 300
|
| 240 |
},
|
| 241 |
{
|
| 242 |
-
"epoch": 0.
|
| 243 |
-
"eval_loss": 0.
|
| 244 |
-
"eval_na_accuracy": 0.
|
| 245 |
-
"eval_ordinal_accuracy": 0.
|
| 246 |
-
"eval_runtime":
|
| 247 |
-
"eval_samples_per_second": 25.
|
| 248 |
-
"eval_steps_per_second": 3.
|
| 249 |
"step": 300
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 0.
|
| 253 |
-
"grad_norm": 0.
|
| 254 |
-
"learning_rate": 0.
|
| 255 |
-
"loss": 0.
|
| 256 |
"step": 310
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"epoch":
|
| 260 |
-
"grad_norm":
|
| 261 |
-
"learning_rate": 0.
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 320
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch":
|
| 267 |
-
"grad_norm":
|
| 268 |
-
"learning_rate": 0.
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 330
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch":
|
| 274 |
-
"grad_norm": 1.
|
| 275 |
-
"learning_rate": 0.
|
| 276 |
-
"loss": 0.
|
| 277 |
"step": 340
|
| 278 |
},
|
| 279 |
{
|
| 280 |
-
"epoch":
|
| 281 |
-
"grad_norm": 0.
|
| 282 |
-
"learning_rate": 0.
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 350
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch":
|
| 288 |
-
"grad_norm": 0.
|
| 289 |
-
"learning_rate": 0.
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 360
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 0.
|
| 297 |
-
"loss": 0.
|
| 298 |
"step": 370
|
| 299 |
},
|
| 300 |
{
|
| 301 |
-
"epoch":
|
| 302 |
-
"grad_norm": 1.
|
| 303 |
-
"learning_rate": 0.
|
| 304 |
-
"loss": 0.
|
| 305 |
"step": 380
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch":
|
| 309 |
-
"grad_norm": 0.
|
| 310 |
-
"learning_rate": 0.
|
| 311 |
-
"loss": 0.
|
| 312 |
"step": 390
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"epoch":
|
| 316 |
-
"grad_norm": 0.
|
| 317 |
-
"learning_rate": 0.
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 400
|
| 320 |
},
|
| 321 |
{
|
| 322 |
-
"epoch":
|
| 323 |
-
"eval_loss": 0.
|
| 324 |
-
"eval_na_accuracy": 0.
|
| 325 |
-
"eval_ordinal_accuracy": 0.
|
| 326 |
-
"eval_runtime":
|
| 327 |
-
"eval_samples_per_second": 25.
|
| 328 |
-
"eval_steps_per_second": 3.
|
| 329 |
"step": 400
|
| 330 |
},
|
| 331 |
{
|
| 332 |
-
"epoch":
|
| 333 |
-
"grad_norm":
|
| 334 |
-
"learning_rate": 0.
|
| 335 |
-
"loss": 0.
|
| 336 |
"step": 410
|
| 337 |
},
|
| 338 |
{
|
| 339 |
-
"epoch":
|
| 340 |
-
"grad_norm":
|
| 341 |
-
"learning_rate": 0.
|
| 342 |
-
"loss": 0.
|
| 343 |
"step": 420
|
| 344 |
},
|
| 345 |
{
|
| 346 |
-
"epoch":
|
| 347 |
-
"grad_norm": 0.
|
| 348 |
-
"learning_rate": 0.
|
| 349 |
-
"loss": 0.
|
| 350 |
"step": 430
|
| 351 |
},
|
| 352 |
{
|
| 353 |
-
"epoch":
|
| 354 |
-
"grad_norm":
|
| 355 |
-
"learning_rate": 0.
|
| 356 |
-
"loss": 0.
|
| 357 |
"step": 440
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch":
|
| 361 |
-
"grad_norm":
|
| 362 |
-
"learning_rate": 0.
|
| 363 |
-
"loss": 0.
|
| 364 |
"step": 450
|
| 365 |
},
|
| 366 |
{
|
| 367 |
-
"epoch":
|
| 368 |
-
"grad_norm":
|
| 369 |
-
"learning_rate": 0.
|
| 370 |
-
"loss": 0.
|
| 371 |
"step": 460
|
| 372 |
},
|
| 373 |
{
|
| 374 |
-
"epoch":
|
| 375 |
-
"grad_norm": 0.
|
| 376 |
-
"learning_rate": 0.
|
| 377 |
-
"loss": 0.
|
| 378 |
"step": 470
|
| 379 |
},
|
| 380 |
{
|
| 381 |
-
"epoch":
|
| 382 |
-
"grad_norm": 0.
|
| 383 |
-
"learning_rate": 0.
|
| 384 |
-
"loss": 0.
|
| 385 |
"step": 480
|
| 386 |
},
|
| 387 |
{
|
| 388 |
-
"epoch":
|
| 389 |
-
"grad_norm":
|
| 390 |
-
"learning_rate": 0.
|
| 391 |
-
"loss": 0.
|
| 392 |
"step": 490
|
| 393 |
},
|
| 394 |
{
|
| 395 |
-
"epoch":
|
| 396 |
-
"grad_norm":
|
| 397 |
-
"learning_rate": 0.
|
| 398 |
-
"loss": 0.
|
| 399 |
"step": 500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch":
|
| 403 |
-
"eval_loss": 0.
|
| 404 |
-
"eval_na_accuracy": 0.
|
| 405 |
-
"eval_ordinal_accuracy": 0.
|
| 406 |
-
"eval_runtime":
|
| 407 |
-
"eval_samples_per_second":
|
| 408 |
-
"eval_steps_per_second": 3.
|
| 409 |
"step": 500
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch":
|
| 413 |
-
"grad_norm":
|
| 414 |
-
"learning_rate": 0.
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 510
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
-
"learning_rate": 0.
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 520
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 0.
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 530
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch":
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate": 0.
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 540
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate": 0.
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 550
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch":
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate": 0.
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 560
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch":
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate": 0.
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 570
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
-
"learning_rate": 0.
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 580
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate": 0.
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 590
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
-
"learning_rate": 0.
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"eval_loss": 0.
|
| 484 |
-
"eval_na_accuracy": 0.
|
| 485 |
-
"eval_ordinal_accuracy": 0.
|
| 486 |
-
"eval_runtime":
|
| 487 |
-
"eval_samples_per_second":
|
| 488 |
-
"eval_steps_per_second": 3.
|
| 489 |
"step": 600
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch":
|
| 493 |
-
"grad_norm":
|
| 494 |
-
"learning_rate": 0.
|
| 495 |
-
"loss": 0.
|
| 496 |
"step": 610
|
| 497 |
},
|
| 498 |
{
|
| 499 |
-
"epoch":
|
| 500 |
-
"grad_norm":
|
| 501 |
-
"learning_rate": 0.
|
| 502 |
-
"loss": 0.
|
| 503 |
"step": 620
|
| 504 |
},
|
| 505 |
{
|
| 506 |
-
"epoch":
|
| 507 |
-
"grad_norm": 0.
|
| 508 |
-
"learning_rate":
|
| 509 |
-
"loss": 0.
|
| 510 |
"step": 630
|
| 511 |
},
|
| 512 |
{
|
| 513 |
-
"epoch":
|
| 514 |
-
"grad_norm": 0.
|
| 515 |
-
"learning_rate":
|
| 516 |
-
"loss": 0.
|
| 517 |
"step": 640
|
| 518 |
},
|
| 519 |
{
|
| 520 |
-
"epoch":
|
| 521 |
-
"grad_norm": 0.
|
| 522 |
-
"learning_rate":
|
| 523 |
-
"loss": 0.
|
| 524 |
"step": 650
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch":
|
| 528 |
-
"grad_norm": 0.
|
| 529 |
-
"learning_rate":
|
| 530 |
-
"loss": 0.
|
| 531 |
"step": 660
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch":
|
| 535 |
-
"grad_norm": 0.
|
| 536 |
-
"learning_rate":
|
| 537 |
-
"loss": 0.
|
| 538 |
"step": 670
|
| 539 |
},
|
| 540 |
{
|
| 541 |
-
"epoch":
|
| 542 |
-
"grad_norm":
|
| 543 |
-
"learning_rate":
|
| 544 |
-
"loss": 0.
|
| 545 |
"step": 680
|
| 546 |
},
|
| 547 |
{
|
| 548 |
-
"epoch":
|
| 549 |
-
"grad_norm": 0.
|
| 550 |
-
"learning_rate":
|
| 551 |
-
"loss": 0.
|
| 552 |
"step": 690
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"epoch":
|
| 556 |
-
"grad_norm":
|
| 557 |
-
"learning_rate":
|
| 558 |
-
"loss": 0.
|
| 559 |
"step": 700
|
| 560 |
},
|
| 561 |
{
|
| 562 |
-
"epoch":
|
| 563 |
-
"eval_loss": 0.
|
| 564 |
-
"eval_na_accuracy": 0.
|
| 565 |
-
"eval_ordinal_accuracy": 0.
|
| 566 |
-
"eval_runtime":
|
| 567 |
-
"eval_samples_per_second":
|
| 568 |
-
"eval_steps_per_second": 3.
|
| 569 |
"step": 700
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch":
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
-
"learning_rate":
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 710
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch":
|
| 580 |
-
"grad_norm":
|
| 581 |
-
"learning_rate":
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 720
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch":
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
-
"learning_rate":
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 730
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 740
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
-
"learning_rate":
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 750
|
| 605 |
},
|
| 606 |
{
|
| 607 |
-
"epoch":
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
-
"learning_rate":
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 760
|
| 612 |
},
|
| 613 |
{
|
| 614 |
-
"epoch":
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
-
"learning_rate":
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 770
|
| 619 |
},
|
| 620 |
{
|
| 621 |
-
"epoch":
|
| 622 |
-
"grad_norm":
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 780
|
| 626 |
},
|
| 627 |
{
|
| 628 |
-
"epoch":
|
| 629 |
-
"grad_norm":
|
| 630 |
-
"learning_rate":
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 790
|
| 633 |
},
|
| 634 |
{
|
| 635 |
-
"epoch":
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
-
"learning_rate":
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 800
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch":
|
| 643 |
-
"eval_loss": 0.
|
| 644 |
-
"eval_na_accuracy": 0.
|
| 645 |
-
"eval_ordinal_accuracy": 0.
|
| 646 |
-
"eval_runtime":
|
| 647 |
-
"eval_samples_per_second": 25.
|
| 648 |
-
"eval_steps_per_second": 3.
|
| 649 |
"step": 800
|
| 650 |
},
|
| 651 |
{
|
| 652 |
-
"epoch":
|
| 653 |
-
"grad_norm":
|
| 654 |
-
"learning_rate":
|
| 655 |
-
"loss": 0.
|
| 656 |
"step": 810
|
| 657 |
},
|
| 658 |
{
|
| 659 |
-
"epoch":
|
| 660 |
-
"grad_norm":
|
| 661 |
-
"learning_rate":
|
| 662 |
-
"loss": 0.
|
| 663 |
"step": 820
|
| 664 |
},
|
| 665 |
{
|
| 666 |
-
"epoch":
|
| 667 |
-
"grad_norm":
|
| 668 |
-
"learning_rate":
|
| 669 |
-
"loss": 0.
|
| 670 |
"step": 830
|
| 671 |
},
|
| 672 |
{
|
| 673 |
-
"epoch":
|
| 674 |
-
"grad_norm":
|
| 675 |
-
"learning_rate":
|
| 676 |
-
"loss": 0.
|
| 677 |
"step": 840
|
| 678 |
},
|
| 679 |
{
|
| 680 |
-
"epoch":
|
| 681 |
-
"grad_norm": 0.
|
| 682 |
-
"learning_rate":
|
| 683 |
-
"loss": 0.
|
| 684 |
"step": 850
|
| 685 |
},
|
| 686 |
{
|
| 687 |
-
"epoch":
|
| 688 |
-
"grad_norm": 0.
|
| 689 |
-
"learning_rate":
|
| 690 |
-
"loss": 0.
|
| 691 |
"step": 860
|
| 692 |
},
|
| 693 |
{
|
| 694 |
-
"epoch":
|
| 695 |
-
"grad_norm":
|
| 696 |
-
"learning_rate":
|
| 697 |
-
"loss": 0.
|
| 698 |
"step": 870
|
| 699 |
},
|
| 700 |
{
|
| 701 |
-
"epoch":
|
| 702 |
-
"grad_norm":
|
| 703 |
-
"learning_rate":
|
| 704 |
-
"loss": 0.
|
| 705 |
"step": 880
|
| 706 |
},
|
| 707 |
{
|
| 708 |
-
"epoch":
|
| 709 |
-
"grad_norm": 0.
|
| 710 |
-
"learning_rate":
|
| 711 |
-
"loss": 0.
|
| 712 |
"step": 890
|
| 713 |
},
|
| 714 |
{
|
| 715 |
-
"epoch":
|
| 716 |
-
"grad_norm":
|
| 717 |
-
"learning_rate":
|
| 718 |
-
"loss": 0.
|
| 719 |
"step": 900
|
| 720 |
},
|
| 721 |
{
|
| 722 |
-
"epoch":
|
| 723 |
-
"eval_loss": 0.
|
| 724 |
-
"eval_na_accuracy": 0.
|
| 725 |
-
"eval_ordinal_accuracy": 0.
|
| 726 |
-
"eval_runtime":
|
| 727 |
-
"eval_samples_per_second":
|
| 728 |
-
"eval_steps_per_second": 3.
|
| 729 |
"step": 900
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
-
"grad_norm":
|
| 734 |
-
"learning_rate":
|
| 735 |
-
"loss": 0.
|
| 736 |
"step": 910
|
| 737 |
},
|
| 738 |
{
|
| 739 |
-
"epoch":
|
| 740 |
-
"grad_norm":
|
| 741 |
-
"learning_rate":
|
| 742 |
-
"loss": 0.
|
| 743 |
"step": 920
|
| 744 |
},
|
| 745 |
{
|
| 746 |
-
"epoch":
|
| 747 |
-
"grad_norm": 0.
|
| 748 |
-
"learning_rate":
|
| 749 |
-
"loss": 0.
|
| 750 |
"step": 930
|
| 751 |
},
|
| 752 |
{
|
| 753 |
-
"epoch":
|
| 754 |
-
"grad_norm": 0.
|
| 755 |
-
"learning_rate":
|
| 756 |
-
"loss": 0.
|
| 757 |
"step": 940
|
| 758 |
},
|
| 759 |
{
|
| 760 |
-
"epoch":
|
| 761 |
-
"grad_norm":
|
| 762 |
-
"learning_rate":
|
| 763 |
-
"loss": 0.
|
| 764 |
"step": 950
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"epoch":
|
| 768 |
-
"grad_norm":
|
| 769 |
-
"learning_rate":
|
| 770 |
-
"loss": 0.
|
| 771 |
"step": 960
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch":
|
| 775 |
-
"grad_norm":
|
| 776 |
-
"learning_rate":
|
| 777 |
-
"loss": 0.
|
| 778 |
"step": 970
|
| 779 |
},
|
| 780 |
{
|
| 781 |
-
"epoch":
|
| 782 |
-
"grad_norm":
|
| 783 |
-
"learning_rate":
|
| 784 |
-
"loss": 0.
|
| 785 |
"step": 980
|
| 786 |
},
|
| 787 |
{
|
| 788 |
-
"epoch":
|
| 789 |
-
"grad_norm": 0.
|
| 790 |
-
"learning_rate":
|
| 791 |
-
"loss": 0.
|
| 792 |
"step": 990
|
| 793 |
},
|
| 794 |
{
|
| 795 |
-
"epoch":
|
| 796 |
-
"grad_norm":
|
| 797 |
-
"learning_rate":
|
| 798 |
-
"loss": 0.
|
| 799 |
"step": 1000
|
| 800 |
},
|
| 801 |
{
|
| 802 |
-
"epoch":
|
| 803 |
-
"eval_loss": 0.
|
| 804 |
-
"eval_na_accuracy": 0.
|
| 805 |
-
"eval_ordinal_accuracy": 0.
|
| 806 |
-
"eval_runtime":
|
| 807 |
-
"eval_samples_per_second":
|
| 808 |
-
"eval_steps_per_second": 3.
|
| 809 |
"step": 1000
|
| 810 |
},
|
| 811 |
{
|
| 812 |
-
"epoch":
|
| 813 |
-
"grad_norm":
|
| 814 |
-
"learning_rate":
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 1010
|
| 817 |
},
|
| 818 |
{
|
| 819 |
-
"epoch":
|
| 820 |
-
"grad_norm":
|
| 821 |
-
"learning_rate":
|
| 822 |
-
"loss": 0.
|
| 823 |
"step": 1020
|
| 824 |
},
|
| 825 |
{
|
| 826 |
-
"epoch":
|
| 827 |
-
"grad_norm": 0.
|
| 828 |
-
"learning_rate":
|
| 829 |
-
"loss": 0.
|
| 830 |
"step": 1030
|
| 831 |
},
|
| 832 |
{
|
| 833 |
-
"epoch":
|
| 834 |
-
"grad_norm": 0.
|
| 835 |
-
"learning_rate":
|
| 836 |
-
"loss": 0.
|
| 837 |
"step": 1040
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch":
|
| 841 |
-
"grad_norm": 0.
|
| 842 |
-
"learning_rate":
|
| 843 |
-
"loss": 0.
|
| 844 |
"step": 1050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
-
"epoch":
|
| 848 |
-
"grad_norm": 0.
|
| 849 |
-
"learning_rate":
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 1060
|
| 852 |
},
|
| 853 |
{
|
| 854 |
-
"epoch":
|
| 855 |
-
"grad_norm":
|
| 856 |
-
"learning_rate":
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 1070
|
| 859 |
},
|
| 860 |
{
|
| 861 |
-
"epoch":
|
| 862 |
-
"grad_norm": 0.
|
| 863 |
-
"learning_rate":
|
| 864 |
-
"loss": 0.
|
| 865 |
"step": 1080
|
| 866 |
},
|
| 867 |
{
|
| 868 |
-
"epoch":
|
| 869 |
-
"grad_norm":
|
| 870 |
-
"learning_rate":
|
| 871 |
-
"loss": 0.
|
| 872 |
"step": 1090
|
| 873 |
},
|
| 874 |
{
|
| 875 |
-
"epoch":
|
| 876 |
-
"grad_norm": 0.
|
| 877 |
-
"learning_rate":
|
| 878 |
-
"loss": 0.
|
| 879 |
"step": 1100
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch":
|
| 883 |
-
"eval_loss": 0.
|
| 884 |
-
"eval_na_accuracy": 0.
|
| 885 |
-
"eval_ordinal_accuracy": 0.
|
| 886 |
-
"eval_runtime":
|
| 887 |
-
"eval_samples_per_second": 25.
|
| 888 |
-
"eval_steps_per_second": 3.
|
| 889 |
"step": 1100
|
| 890 |
},
|
| 891 |
{
|
| 892 |
-
"epoch":
|
| 893 |
-
"grad_norm":
|
| 894 |
-
"learning_rate":
|
| 895 |
-
"loss": 0.
|
| 896 |
"step": 1110
|
| 897 |
},
|
| 898 |
{
|
| 899 |
-
"epoch":
|
| 900 |
-
"grad_norm":
|
| 901 |
-
"learning_rate":
|
| 902 |
-
"loss": 0.
|
| 903 |
"step": 1120
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch":
|
| 907 |
-
"grad_norm":
|
| 908 |
-
"learning_rate":
|
| 909 |
-
"loss": 0.
|
| 910 |
"step": 1130
|
| 911 |
},
|
| 912 |
{
|
| 913 |
-
"epoch":
|
| 914 |
-
"grad_norm": 0.
|
| 915 |
-
"learning_rate":
|
| 916 |
-
"loss": 0.
|
| 917 |
"step": 1140
|
| 918 |
},
|
| 919 |
{
|
| 920 |
-
"epoch":
|
| 921 |
-
"grad_norm": 0.
|
| 922 |
-
"learning_rate":
|
| 923 |
-
"loss": 0.
|
| 924 |
"step": 1150
|
| 925 |
},
|
| 926 |
{
|
| 927 |
-
"epoch":
|
| 928 |
-
"grad_norm": 0.
|
| 929 |
-
"learning_rate":
|
| 930 |
-
"loss": 0.
|
| 931 |
"step": 1160
|
| 932 |
},
|
| 933 |
{
|
| 934 |
-
"epoch":
|
| 935 |
-
"grad_norm": 0.
|
| 936 |
-
"learning_rate":
|
| 937 |
-
"loss": 0.
|
| 938 |
"step": 1170
|
| 939 |
},
|
| 940 |
{
|
| 941 |
-
"epoch":
|
| 942 |
-
"grad_norm": 0.
|
| 943 |
-
"learning_rate":
|
| 944 |
-
"loss": 0.
|
| 945 |
"step": 1180
|
| 946 |
},
|
| 947 |
{
|
| 948 |
-
"epoch":
|
| 949 |
-
"grad_norm": 0.
|
| 950 |
-
"learning_rate":
|
| 951 |
-
"loss": 0.
|
| 952 |
"step": 1190
|
| 953 |
},
|
| 954 |
{
|
| 955 |
-
"epoch":
|
| 956 |
-
"grad_norm":
|
| 957 |
-
"learning_rate":
|
| 958 |
-
"loss": 0.
|
| 959 |
"step": 1200
|
| 960 |
},
|
| 961 |
{
|
| 962 |
-
"epoch":
|
| 963 |
-
"eval_loss": 0.
|
| 964 |
-
"eval_na_accuracy": 0.
|
| 965 |
-
"eval_ordinal_accuracy": 0.
|
| 966 |
-
"eval_runtime":
|
| 967 |
-
"eval_samples_per_second":
|
| 968 |
-
"eval_steps_per_second": 3.
|
| 969 |
"step": 1200
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch":
|
| 973 |
-
"grad_norm": 0.
|
| 974 |
-
"learning_rate":
|
| 975 |
-
"loss": 0.
|
| 976 |
"step": 1210
|
| 977 |
},
|
| 978 |
{
|
| 979 |
-
"epoch":
|
| 980 |
-
"grad_norm": 0.
|
| 981 |
-
"learning_rate":
|
| 982 |
-
"loss": 0.
|
| 983 |
"step": 1220
|
| 984 |
},
|
| 985 |
{
|
| 986 |
-
"epoch":
|
| 987 |
-
"grad_norm":
|
| 988 |
-
"learning_rate":
|
| 989 |
-
"loss": 0.
|
| 990 |
"step": 1230
|
| 991 |
},
|
| 992 |
{
|
| 993 |
-
"epoch":
|
| 994 |
-
"grad_norm": 0.
|
| 995 |
-
"learning_rate":
|
| 996 |
-
"loss": 0.
|
| 997 |
"step": 1240
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
-
"epoch":
|
| 1001 |
-
"grad_norm": 0.
|
| 1002 |
-
"learning_rate":
|
| 1003 |
-
"loss": 0.
|
| 1004 |
"step": 1250
|
| 1005 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1006 |
{
|
| 1007 |
"epoch": 4.0,
|
| 1008 |
-
"step":
|
| 1009 |
-
"total_flos":
|
| 1010 |
-
"train_loss": 0.
|
| 1011 |
-
"train_runtime":
|
| 1012 |
-
"train_samples_per_second":
|
| 1013 |
-
"train_steps_per_second": 0.
|
| 1014 |
}
|
| 1015 |
],
|
| 1016 |
"logging_steps": 10,
|
| 1017 |
-
"max_steps":
|
| 1018 |
"num_input_tokens_seen": 0,
|
| 1019 |
"num_train_epochs": 4,
|
| 1020 |
"save_steps": 100,
|
| 1021 |
-
"total_flos":
|
| 1022 |
"train_batch_size": 16,
|
| 1023 |
"trial_name": null,
|
| 1024 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.2531912922859192,
|
| 3 |
+
"best_model_checkpoint": "./ryan_model314/checkpoint-600",
|
| 4 |
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 2500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 1.1103402376174927,
|
| 14 |
+
"learning_rate": 0.00019920000000000002,
|
| 15 |
+
"loss": 0.5731,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.03,
|
| 20 |
+
"grad_norm": 0.8193413019180298,
|
| 21 |
+
"learning_rate": 0.0001984,
|
| 22 |
+
"loss": 0.4217,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 1.3536686897277832,
|
| 28 |
+
"learning_rate": 0.0001976,
|
| 29 |
+
"loss": 0.3709,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.06,
|
| 34 |
+
"grad_norm": 0.9998810887336731,
|
| 35 |
+
"learning_rate": 0.0001968,
|
| 36 |
+
"loss": 0.3398,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.08,
|
| 41 |
+
"grad_norm": 1.5689244270324707,
|
| 42 |
+
"learning_rate": 0.000196,
|
| 43 |
+
"loss": 0.3346,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.1,
|
| 48 |
+
"grad_norm": 1.1778826713562012,
|
| 49 |
+
"learning_rate": 0.0001952,
|
| 50 |
+
"loss": 0.3406,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.11,
|
| 55 |
+
"grad_norm": 1.3193926811218262,
|
| 56 |
+
"learning_rate": 0.0001944,
|
| 57 |
+
"loss": 0.2755,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.13,
|
| 62 |
+
"grad_norm": 1.1302804946899414,
|
| 63 |
+
"learning_rate": 0.00019360000000000002,
|
| 64 |
+
"loss": 0.3944,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.14,
|
| 69 |
+
"grad_norm": 0.8255844712257385,
|
| 70 |
+
"learning_rate": 0.0001928,
|
| 71 |
+
"loss": 0.3473,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.16,
|
| 76 |
+
"grad_norm": 1.0871790647506714,
|
| 77 |
+
"learning_rate": 0.000192,
|
| 78 |
+
"loss": 0.3042,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.16,
|
| 83 |
+
"eval_loss": 0.3673088252544403,
|
| 84 |
+
"eval_na_accuracy": 0.928,
|
| 85 |
+
"eval_ordinal_accuracy": 0.4671280276816609,
|
| 86 |
+
"eval_runtime": 110.8646,
|
| 87 |
+
"eval_samples_per_second": 9.02,
|
| 88 |
+
"eval_steps_per_second": 1.128,
|
| 89 |
"step": 100
|
| 90 |
},
|
| 91 |
{
|
| 92 |
+
"epoch": 0.18,
|
| 93 |
+
"grad_norm": 1.5816177129745483,
|
| 94 |
+
"learning_rate": 0.0001912,
|
| 95 |
+
"loss": 0.4058,
|
| 96 |
"step": 110
|
| 97 |
},
|
| 98 |
{
|
| 99 |
+
"epoch": 0.19,
|
| 100 |
+
"grad_norm": 0.8431822061538696,
|
| 101 |
+
"learning_rate": 0.0001904,
|
| 102 |
+
"loss": 0.2781,
|
| 103 |
"step": 120
|
| 104 |
},
|
| 105 |
{
|
| 106 |
+
"epoch": 0.21,
|
| 107 |
+
"grad_norm": 1.0826754570007324,
|
| 108 |
+
"learning_rate": 0.0001896,
|
| 109 |
+
"loss": 0.2587,
|
| 110 |
"step": 130
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"grad_norm": 3.7366294860839844,
|
| 115 |
+
"learning_rate": 0.0001888,
|
| 116 |
+
"loss": 0.3432,
|
| 117 |
"step": 140
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.24,
|
| 121 |
+
"grad_norm": 0.5233088731765747,
|
| 122 |
+
"learning_rate": 0.000188,
|
| 123 |
+
"loss": 0.3484,
|
| 124 |
"step": 150
|
| 125 |
},
|
| 126 |
{
|
| 127 |
+
"epoch": 0.26,
|
| 128 |
+
"grad_norm": 1.8766111135482788,
|
| 129 |
+
"learning_rate": 0.00018720000000000002,
|
| 130 |
+
"loss": 0.3597,
|
| 131 |
"step": 160
|
| 132 |
},
|
| 133 |
{
|
| 134 |
+
"epoch": 0.27,
|
| 135 |
+
"grad_norm": 1.0037935972213745,
|
| 136 |
+
"learning_rate": 0.00018640000000000003,
|
| 137 |
+
"loss": 0.288,
|
| 138 |
"step": 170
|
| 139 |
},
|
| 140 |
{
|
| 141 |
+
"epoch": 0.29,
|
| 142 |
+
"grad_norm": 1.3281046152114868,
|
| 143 |
+
"learning_rate": 0.0001856,
|
| 144 |
+
"loss": 0.3207,
|
| 145 |
"step": 180
|
| 146 |
},
|
| 147 |
{
|
| 148 |
+
"epoch": 0.3,
|
| 149 |
+
"grad_norm": 1.4793013334274292,
|
| 150 |
+
"learning_rate": 0.00018480000000000002,
|
| 151 |
+
"loss": 0.3372,
|
| 152 |
"step": 190
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.32,
|
| 156 |
+
"grad_norm": 0.8796727657318115,
|
| 157 |
+
"learning_rate": 0.00018400000000000003,
|
| 158 |
+
"loss": 0.2904,
|
| 159 |
"step": 200
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.32,
|
| 163 |
+
"eval_loss": 0.29769936203956604,
|
| 164 |
+
"eval_na_accuracy": 0.933,
|
| 165 |
+
"eval_ordinal_accuracy": 0.5790080738177624,
|
| 166 |
+
"eval_runtime": 39.6284,
|
| 167 |
+
"eval_samples_per_second": 25.234,
|
| 168 |
+
"eval_steps_per_second": 3.154,
|
| 169 |
"step": 200
|
| 170 |
},
|
| 171 |
{
|
| 172 |
+
"epoch": 0.34,
|
| 173 |
+
"grad_norm": 1.3502057790756226,
|
| 174 |
+
"learning_rate": 0.0001832,
|
| 175 |
+
"loss": 0.3519,
|
| 176 |
"step": 210
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.35,
|
| 180 |
+
"grad_norm": 1.5546174049377441,
|
| 181 |
+
"learning_rate": 0.00018240000000000002,
|
| 182 |
+
"loss": 0.3243,
|
| 183 |
"step": 220
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.37,
|
| 187 |
+
"grad_norm": 0.7677227854728699,
|
| 188 |
+
"learning_rate": 0.00018160000000000002,
|
| 189 |
+
"loss": 0.2914,
|
| 190 |
"step": 230
|
| 191 |
},
|
| 192 |
{
|
| 193 |
+
"epoch": 0.38,
|
| 194 |
+
"grad_norm": 1.1754639148712158,
|
| 195 |
+
"learning_rate": 0.0001808,
|
| 196 |
+
"loss": 0.3539,
|
| 197 |
"step": 240
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 0.4,
|
| 201 |
+
"grad_norm": 0.8472470641136169,
|
| 202 |
+
"learning_rate": 0.00018,
|
| 203 |
+
"loss": 0.2395,
|
| 204 |
"step": 250
|
| 205 |
},
|
| 206 |
{
|
| 207 |
+
"epoch": 0.42,
|
| 208 |
+
"grad_norm": 1.1917964220046997,
|
| 209 |
+
"learning_rate": 0.00017920000000000002,
|
| 210 |
+
"loss": 0.2295,
|
| 211 |
"step": 260
|
| 212 |
},
|
| 213 |
{
|
| 214 |
+
"epoch": 0.43,
|
| 215 |
+
"grad_norm": 0.7398644685745239,
|
| 216 |
+
"learning_rate": 0.0001784,
|
| 217 |
+
"loss": 0.2398,
|
| 218 |
"step": 270
|
| 219 |
},
|
| 220 |
{
|
| 221 |
+
"epoch": 0.45,
|
| 222 |
+
"grad_norm": 0.5953208804130554,
|
| 223 |
+
"learning_rate": 0.0001776,
|
| 224 |
+
"loss": 0.2786,
|
| 225 |
"step": 280
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 0.46,
|
| 229 |
+
"grad_norm": 2.0648913383483887,
|
| 230 |
+
"learning_rate": 0.00017680000000000001,
|
| 231 |
+
"loss": 0.3661,
|
| 232 |
"step": 290
|
| 233 |
},
|
| 234 |
{
|
| 235 |
+
"epoch": 0.48,
|
| 236 |
+
"grad_norm": 1.4048805236816406,
|
| 237 |
+
"learning_rate": 0.00017600000000000002,
|
| 238 |
+
"loss": 0.2648,
|
| 239 |
"step": 300
|
| 240 |
},
|
| 241 |
{
|
| 242 |
+
"epoch": 0.48,
|
| 243 |
+
"eval_loss": 0.2830840051174164,
|
| 244 |
+
"eval_na_accuracy": 0.944,
|
| 245 |
+
"eval_ordinal_accuracy": 0.5940023068050749,
|
| 246 |
+
"eval_runtime": 39.7255,
|
| 247 |
+
"eval_samples_per_second": 25.173,
|
| 248 |
+
"eval_steps_per_second": 3.147,
|
| 249 |
"step": 300
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 0.5,
|
| 253 |
+
"grad_norm": 0.8102580904960632,
|
| 254 |
+
"learning_rate": 0.0001752,
|
| 255 |
+
"loss": 0.2359,
|
| 256 |
"step": 310
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"epoch": 0.51,
|
| 260 |
+
"grad_norm": 2.0220913887023926,
|
| 261 |
+
"learning_rate": 0.0001744,
|
| 262 |
+
"loss": 0.2557,
|
| 263 |
"step": 320
|
| 264 |
},
|
| 265 |
{
|
| 266 |
+
"epoch": 0.53,
|
| 267 |
+
"grad_norm": 1.2111886739730835,
|
| 268 |
+
"learning_rate": 0.00017360000000000002,
|
| 269 |
+
"loss": 0.3025,
|
| 270 |
"step": 330
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 0.54,
|
| 274 |
+
"grad_norm": 1.788378119468689,
|
| 275 |
+
"learning_rate": 0.0001728,
|
| 276 |
+
"loss": 0.3067,
|
| 277 |
"step": 340
|
| 278 |
},
|
| 279 |
{
|
| 280 |
+
"epoch": 0.56,
|
| 281 |
+
"grad_norm": 0.7332974076271057,
|
| 282 |
+
"learning_rate": 0.000172,
|
| 283 |
+
"loss": 0.2612,
|
| 284 |
"step": 350
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 0.58,
|
| 288 |
+
"grad_norm": 0.5220205783843994,
|
| 289 |
+
"learning_rate": 0.00017120000000000001,
|
| 290 |
+
"loss": 0.2924,
|
| 291 |
"step": 360
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 0.59,
|
| 295 |
+
"grad_norm": 0.8991191387176514,
|
| 296 |
+
"learning_rate": 0.0001704,
|
| 297 |
+
"loss": 0.2379,
|
| 298 |
"step": 370
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 0.61,
|
| 302 |
+
"grad_norm": 1.6633837223052979,
|
| 303 |
+
"learning_rate": 0.0001696,
|
| 304 |
+
"loss": 0.2792,
|
| 305 |
"step": 380
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 0.62,
|
| 309 |
+
"grad_norm": 0.9553330540657043,
|
| 310 |
+
"learning_rate": 0.0001688,
|
| 311 |
+
"loss": 0.2512,
|
| 312 |
"step": 390
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"epoch": 0.64,
|
| 316 |
+
"grad_norm": 0.4544942080974579,
|
| 317 |
+
"learning_rate": 0.000168,
|
| 318 |
+
"loss": 0.3036,
|
| 319 |
"step": 400
|
| 320 |
},
|
| 321 |
{
|
| 322 |
+
"epoch": 0.64,
|
| 323 |
+
"eval_loss": 0.27759096026420593,
|
| 324 |
+
"eval_na_accuracy": 0.949,
|
| 325 |
+
"eval_ordinal_accuracy": 0.5870818915801614,
|
| 326 |
+
"eval_runtime": 39.7611,
|
| 327 |
+
"eval_samples_per_second": 25.15,
|
| 328 |
+
"eval_steps_per_second": 3.144,
|
| 329 |
"step": 400
|
| 330 |
},
|
| 331 |
{
|
| 332 |
+
"epoch": 0.66,
|
| 333 |
+
"grad_norm": 1.2751814126968384,
|
| 334 |
+
"learning_rate": 0.0001672,
|
| 335 |
+
"loss": 0.3042,
|
| 336 |
"step": 410
|
| 337 |
},
|
| 338 |
{
|
| 339 |
+
"epoch": 0.67,
|
| 340 |
+
"grad_norm": 1.791074514389038,
|
| 341 |
+
"learning_rate": 0.0001664,
|
| 342 |
+
"loss": 0.3341,
|
| 343 |
"step": 420
|
| 344 |
},
|
| 345 |
{
|
| 346 |
+
"epoch": 0.69,
|
| 347 |
+
"grad_norm": 0.9887642860412598,
|
| 348 |
+
"learning_rate": 0.0001656,
|
| 349 |
+
"loss": 0.2868,
|
| 350 |
"step": 430
|
| 351 |
},
|
| 352 |
{
|
| 353 |
+
"epoch": 0.7,
|
| 354 |
+
"grad_norm": 1.3511923551559448,
|
| 355 |
+
"learning_rate": 0.0001648,
|
| 356 |
+
"loss": 0.3763,
|
| 357 |
"step": 440
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 0.72,
|
| 361 |
+
"grad_norm": 1.7992609739303589,
|
| 362 |
+
"learning_rate": 0.000164,
|
| 363 |
+
"loss": 0.2264,
|
| 364 |
"step": 450
|
| 365 |
},
|
| 366 |
{
|
| 367 |
+
"epoch": 0.74,
|
| 368 |
+
"grad_norm": 1.0241813659667969,
|
| 369 |
+
"learning_rate": 0.0001632,
|
| 370 |
+
"loss": 0.3018,
|
| 371 |
"step": 460
|
| 372 |
},
|
| 373 |
{
|
| 374 |
+
"epoch": 0.75,
|
| 375 |
+
"grad_norm": 0.628193736076355,
|
| 376 |
+
"learning_rate": 0.00016240000000000002,
|
| 377 |
+
"loss": 0.3323,
|
| 378 |
"step": 470
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 0.77,
|
| 382 |
+
"grad_norm": 0.8471026420593262,
|
| 383 |
+
"learning_rate": 0.00016160000000000002,
|
| 384 |
+
"loss": 0.2005,
|
| 385 |
"step": 480
|
| 386 |
},
|
| 387 |
{
|
| 388 |
+
"epoch": 0.78,
|
| 389 |
+
"grad_norm": 1.1799852848052979,
|
| 390 |
+
"learning_rate": 0.0001608,
|
| 391 |
+
"loss": 0.2984,
|
| 392 |
"step": 490
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 0.8,
|
| 396 |
+
"grad_norm": 1.595058560371399,
|
| 397 |
+
"learning_rate": 0.00016,
|
| 398 |
+
"loss": 0.2656,
|
| 399 |
"step": 500
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 0.8,
|
| 403 |
+
"eval_loss": 0.2846027612686157,
|
| 404 |
+
"eval_na_accuracy": 0.931,
|
| 405 |
+
"eval_ordinal_accuracy": 0.6101499423298731,
|
| 406 |
+
"eval_runtime": 39.9154,
|
| 407 |
+
"eval_samples_per_second": 25.053,
|
| 408 |
+
"eval_steps_per_second": 3.132,
|
| 409 |
"step": 500
|
| 410 |
},
|
| 411 |
{
|
| 412 |
+
"epoch": 0.82,
|
| 413 |
+
"grad_norm": 0.9246352910995483,
|
| 414 |
+
"learning_rate": 0.00015920000000000002,
|
| 415 |
+
"loss": 0.3591,
|
| 416 |
"step": 510
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.83,
|
| 420 |
+
"grad_norm": 0.9456105828285217,
|
| 421 |
+
"learning_rate": 0.00015840000000000003,
|
| 422 |
+
"loss": 0.3569,
|
| 423 |
"step": 520
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 0.85,
|
| 427 |
+
"grad_norm": 1.111274003982544,
|
| 428 |
+
"learning_rate": 0.0001576,
|
| 429 |
+
"loss": 0.3243,
|
| 430 |
"step": 530
|
| 431 |
},
|
| 432 |
{
|
| 433 |
+
"epoch": 0.86,
|
| 434 |
+
"grad_norm": NaN,
|
| 435 |
+
"learning_rate": 0.00015688,
|
| 436 |
+
"loss": 0.2911,
|
| 437 |
"step": 540
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 0.88,
|
| 441 |
+
"grad_norm": 0.8232502341270447,
|
| 442 |
+
"learning_rate": 0.00015616000000000002,
|
| 443 |
+
"loss": 0.3236,
|
| 444 |
"step": 550
|
| 445 |
},
|
| 446 |
{
|
| 447 |
+
"epoch": 0.9,
|
| 448 |
+
"grad_norm": 0.6359846591949463,
|
| 449 |
+
"learning_rate": 0.00015536,
|
| 450 |
+
"loss": 0.3211,
|
| 451 |
"step": 560
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 0.91,
|
| 455 |
+
"grad_norm": 0.545005738735199,
|
| 456 |
+
"learning_rate": 0.00015456,
|
| 457 |
+
"loss": 0.205,
|
| 458 |
"step": 570
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.93,
|
| 462 |
+
"grad_norm": 0.6029797196388245,
|
| 463 |
+
"learning_rate": 0.00015376000000000002,
|
| 464 |
+
"loss": 0.1928,
|
| 465 |
"step": 580
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 0.94,
|
| 469 |
+
"grad_norm": 0.7442355155944824,
|
| 470 |
+
"learning_rate": 0.00015296000000000003,
|
| 471 |
+
"loss": 0.3273,
|
| 472 |
"step": 590
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 0.96,
|
| 476 |
+
"grad_norm": 0.6751519441604614,
|
| 477 |
+
"learning_rate": 0.00015216,
|
| 478 |
+
"loss": 0.2954,
|
| 479 |
"step": 600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 0.96,
|
| 483 |
+
"eval_loss": 0.2531912922859192,
|
| 484 |
+
"eval_na_accuracy": 0.947,
|
| 485 |
+
"eval_ordinal_accuracy": 0.5951557093425606,
|
| 486 |
+
"eval_runtime": 39.4037,
|
| 487 |
+
"eval_samples_per_second": 25.378,
|
| 488 |
+
"eval_steps_per_second": 3.172,
|
| 489 |
"step": 600
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 0.98,
|
| 493 |
+
"grad_norm": 1.6347012519836426,
|
| 494 |
+
"learning_rate": 0.00015136000000000001,
|
| 495 |
+
"loss": 0.2256,
|
| 496 |
"step": 610
|
| 497 |
},
|
| 498 |
{
|
| 499 |
+
"epoch": 0.99,
|
| 500 |
+
"grad_norm": 6.180319309234619,
|
| 501 |
+
"learning_rate": 0.00015056000000000002,
|
| 502 |
+
"loss": 0.2003,
|
| 503 |
"step": 620
|
| 504 |
},
|
| 505 |
{
|
| 506 |
+
"epoch": 1.01,
|
| 507 |
+
"grad_norm": 0.8919633626937866,
|
| 508 |
+
"learning_rate": 0.00014976,
|
| 509 |
+
"loss": 0.198,
|
| 510 |
"step": 630
|
| 511 |
},
|
| 512 |
{
|
| 513 |
+
"epoch": 1.02,
|
| 514 |
+
"grad_norm": 0.9197341203689575,
|
| 515 |
+
"learning_rate": 0.00014896,
|
| 516 |
+
"loss": 0.1685,
|
| 517 |
"step": 640
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 1.04,
|
| 521 |
+
"grad_norm": 0.37014976143836975,
|
| 522 |
+
"learning_rate": 0.00014816000000000002,
|
| 523 |
+
"loss": 0.1729,
|
| 524 |
"step": 650
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 1.06,
|
| 528 |
+
"grad_norm": 0.8919755220413208,
|
| 529 |
+
"learning_rate": 0.00014736,
|
| 530 |
+
"loss": 0.1993,
|
| 531 |
"step": 660
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 1.07,
|
| 535 |
+
"grad_norm": 0.7291600704193115,
|
| 536 |
+
"learning_rate": 0.00014656,
|
| 537 |
+
"loss": 0.1893,
|
| 538 |
"step": 670
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 1.09,
|
| 542 |
+
"grad_norm": 2.347400665283203,
|
| 543 |
+
"learning_rate": 0.00014576000000000001,
|
| 544 |
+
"loss": 0.1799,
|
| 545 |
"step": 680
|
| 546 |
},
|
| 547 |
{
|
| 548 |
+
"epoch": 1.1,
|
| 549 |
+
"grad_norm": 0.3188568949699402,
|
| 550 |
+
"learning_rate": 0.00014496,
|
| 551 |
+
"loss": 0.167,
|
| 552 |
"step": 690
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"epoch": 1.12,
|
| 556 |
+
"grad_norm": 1.342278242111206,
|
| 557 |
+
"learning_rate": 0.00014416,
|
| 558 |
+
"loss": 0.1991,
|
| 559 |
"step": 700
|
| 560 |
},
|
| 561 |
{
|
| 562 |
+
"epoch": 1.12,
|
| 563 |
+
"eval_loss": 0.26034072041511536,
|
| 564 |
+
"eval_na_accuracy": 0.942,
|
| 565 |
+
"eval_ordinal_accuracy": 0.6078431372549019,
|
| 566 |
+
"eval_runtime": 39.5088,
|
| 567 |
+
"eval_samples_per_second": 25.311,
|
| 568 |
+
"eval_steps_per_second": 3.164,
|
| 569 |
"step": 700
|
| 570 |
},
|
| 571 |
{
|
| 572 |
+
"epoch": 1.14,
|
| 573 |
+
"grad_norm": 0.48655757308006287,
|
| 574 |
+
"learning_rate": 0.00014336,
|
| 575 |
+
"loss": 0.1885,
|
| 576 |
"step": 710
|
| 577 |
},
|
| 578 |
{
|
| 579 |
+
"epoch": 1.15,
|
| 580 |
+
"grad_norm": 0.556333065032959,
|
| 581 |
+
"learning_rate": 0.00014256000000000002,
|
| 582 |
+
"loss": 0.1449,
|
| 583 |
"step": 720
|
| 584 |
},
|
| 585 |
{
|
| 586 |
+
"epoch": 1.17,
|
| 587 |
+
"grad_norm": 0.4880894422531128,
|
| 588 |
+
"learning_rate": 0.00014176,
|
| 589 |
+
"loss": 0.1164,
|
| 590 |
"step": 730
|
| 591 |
},
|
| 592 |
{
|
| 593 |
+
"epoch": 1.18,
|
| 594 |
+
"grad_norm": 0.599926233291626,
|
| 595 |
+
"learning_rate": 0.00014096,
|
| 596 |
+
"loss": 0.2113,
|
| 597 |
"step": 740
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 1.2,
|
| 601 |
+
"grad_norm": 0.6070149540901184,
|
| 602 |
+
"learning_rate": 0.00014016,
|
| 603 |
+
"loss": 0.1534,
|
| 604 |
"step": 750
|
| 605 |
},
|
| 606 |
{
|
| 607 |
+
"epoch": 1.22,
|
| 608 |
+
"grad_norm": 0.7789746522903442,
|
| 609 |
+
"learning_rate": 0.00013936,
|
| 610 |
+
"loss": 0.1655,
|
| 611 |
"step": 760
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 1.23,
|
| 615 |
+
"grad_norm": 0.5523375868797302,
|
| 616 |
+
"learning_rate": 0.00013856,
|
| 617 |
+
"loss": 0.298,
|
| 618 |
"step": 770
|
| 619 |
},
|
| 620 |
{
|
| 621 |
+
"epoch": 1.25,
|
| 622 |
+
"grad_norm": 2.4257819652557373,
|
| 623 |
+
"learning_rate": 0.00013776,
|
| 624 |
+
"loss": 0.2101,
|
| 625 |
"step": 780
|
| 626 |
},
|
| 627 |
{
|
| 628 |
+
"epoch": 1.26,
|
| 629 |
+
"grad_norm": 0.5729731321334839,
|
| 630 |
+
"learning_rate": 0.00013696,
|
| 631 |
+
"loss": 0.133,
|
| 632 |
"step": 790
|
| 633 |
},
|
| 634 |
{
|
| 635 |
+
"epoch": 1.28,
|
| 636 |
+
"grad_norm": 0.4050444960594177,
|
| 637 |
+
"learning_rate": 0.00013616,
|
| 638 |
+
"loss": 0.1678,
|
| 639 |
"step": 800
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 1.28,
|
| 643 |
+
"eval_loss": 0.2904650568962097,
|
| 644 |
+
"eval_na_accuracy": 0.942,
|
| 645 |
+
"eval_ordinal_accuracy": 0.6332179930795848,
|
| 646 |
+
"eval_runtime": 39.7139,
|
| 647 |
+
"eval_samples_per_second": 25.18,
|
| 648 |
+
"eval_steps_per_second": 3.148,
|
| 649 |
"step": 800
|
| 650 |
},
|
| 651 |
{
|
| 652 |
+
"epoch": 1.3,
|
| 653 |
+
"grad_norm": 4.782747745513916,
|
| 654 |
+
"learning_rate": 0.00013536,
|
| 655 |
+
"loss": 0.207,
|
| 656 |
"step": 810
|
| 657 |
},
|
| 658 |
{
|
| 659 |
+
"epoch": 1.31,
|
| 660 |
+
"grad_norm": 2.577669143676758,
|
| 661 |
+
"learning_rate": 0.00013455999999999999,
|
| 662 |
+
"loss": 0.1818,
|
| 663 |
"step": 820
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 1.33,
|
| 667 |
+
"grad_norm": 2.8163273334503174,
|
| 668 |
+
"learning_rate": 0.00013376,
|
| 669 |
+
"loss": 0.1761,
|
| 670 |
"step": 830
|
| 671 |
},
|
| 672 |
{
|
| 673 |
+
"epoch": 1.34,
|
| 674 |
+
"grad_norm": 2.213799238204956,
|
| 675 |
+
"learning_rate": 0.00013296,
|
| 676 |
+
"loss": 0.2966,
|
| 677 |
"step": 840
|
| 678 |
},
|
| 679 |
{
|
| 680 |
+
"epoch": 1.36,
|
| 681 |
+
"grad_norm": 0.8946444988250732,
|
| 682 |
+
"learning_rate": 0.00013216,
|
| 683 |
+
"loss": 0.1569,
|
| 684 |
"step": 850
|
| 685 |
},
|
| 686 |
{
|
| 687 |
+
"epoch": 1.38,
|
| 688 |
+
"grad_norm": 0.6494708061218262,
|
| 689 |
+
"learning_rate": 0.00013136000000000002,
|
| 690 |
+
"loss": 0.1746,
|
| 691 |
"step": 860
|
| 692 |
},
|
| 693 |
{
|
| 694 |
+
"epoch": 1.39,
|
| 695 |
+
"grad_norm": 1.0058079957962036,
|
| 696 |
+
"learning_rate": 0.00013056000000000002,
|
| 697 |
+
"loss": 0.1204,
|
| 698 |
"step": 870
|
| 699 |
},
|
| 700 |
{
|
| 701 |
+
"epoch": 1.41,
|
| 702 |
+
"grad_norm": 1.1752161979675293,
|
| 703 |
+
"learning_rate": 0.00012976,
|
| 704 |
+
"loss": 0.2082,
|
| 705 |
"step": 880
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 1.42,
|
| 709 |
+
"grad_norm": 0.5655858516693115,
|
| 710 |
+
"learning_rate": 0.00012896,
|
| 711 |
+
"loss": 0.1971,
|
| 712 |
"step": 890
|
| 713 |
},
|
| 714 |
{
|
| 715 |
+
"epoch": 1.44,
|
| 716 |
+
"grad_norm": 2.5486743450164795,
|
| 717 |
+
"learning_rate": 0.00012816000000000002,
|
| 718 |
+
"loss": 0.2514,
|
| 719 |
"step": 900
|
| 720 |
},
|
| 721 |
{
|
| 722 |
+
"epoch": 1.44,
|
| 723 |
+
"eval_loss": 0.25656750798225403,
|
| 724 |
+
"eval_na_accuracy": 0.94,
|
| 725 |
+
"eval_ordinal_accuracy": 0.6089965397923875,
|
| 726 |
+
"eval_runtime": 39.7194,
|
| 727 |
+
"eval_samples_per_second": 25.177,
|
| 728 |
+
"eval_steps_per_second": 3.147,
|
| 729 |
"step": 900
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 1.46,
|
| 733 |
+
"grad_norm": 0.878511369228363,
|
| 734 |
+
"learning_rate": 0.00012736,
|
| 735 |
+
"loss": 0.162,
|
| 736 |
"step": 910
|
| 737 |
},
|
| 738 |
{
|
| 739 |
+
"epoch": 1.47,
|
| 740 |
+
"grad_norm": 1.1985282897949219,
|
| 741 |
+
"learning_rate": 0.00012656,
|
| 742 |
+
"loss": 0.2268,
|
| 743 |
"step": 920
|
| 744 |
},
|
| 745 |
{
|
| 746 |
+
"epoch": 1.49,
|
| 747 |
+
"grad_norm": 0.521425187587738,
|
| 748 |
+
"learning_rate": 0.00012576000000000002,
|
| 749 |
+
"loss": 0.1556,
|
| 750 |
"step": 930
|
| 751 |
},
|
| 752 |
{
|
| 753 |
+
"epoch": 1.5,
|
| 754 |
+
"grad_norm": 0.9773241877555847,
|
| 755 |
+
"learning_rate": 0.00012496000000000002,
|
| 756 |
+
"loss": 0.1457,
|
| 757 |
"step": 940
|
| 758 |
},
|
| 759 |
{
|
| 760 |
+
"epoch": 1.52,
|
| 761 |
+
"grad_norm": 1.6476322412490845,
|
| 762 |
+
"learning_rate": 0.00012416,
|
| 763 |
+
"loss": 0.1913,
|
| 764 |
"step": 950
|
| 765 |
},
|
| 766 |
{
|
| 767 |
+
"epoch": 1.54,
|
| 768 |
+
"grad_norm": 1.7127236127853394,
|
| 769 |
+
"learning_rate": 0.00012336,
|
| 770 |
+
"loss": 0.1961,
|
| 771 |
"step": 960
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 1.55,
|
| 775 |
+
"grad_norm": 4.41243314743042,
|
| 776 |
+
"learning_rate": 0.00012256000000000002,
|
| 777 |
+
"loss": 0.2061,
|
| 778 |
"step": 970
|
| 779 |
},
|
| 780 |
{
|
| 781 |
+
"epoch": 1.57,
|
| 782 |
+
"grad_norm": 1.5907992124557495,
|
| 783 |
+
"learning_rate": 0.00012176000000000001,
|
| 784 |
+
"loss": 0.1299,
|
| 785 |
"step": 980
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 1.58,
|
| 789 |
+
"grad_norm": 0.5711427927017212,
|
| 790 |
+
"learning_rate": 0.00012096000000000001,
|
| 791 |
+
"loss": 0.1755,
|
| 792 |
"step": 990
|
| 793 |
},
|
| 794 |
{
|
| 795 |
+
"epoch": 1.6,
|
| 796 |
+
"grad_norm": 2.925363302230835,
|
| 797 |
+
"learning_rate": 0.00012016,
|
| 798 |
+
"loss": 0.2328,
|
| 799 |
"step": 1000
|
| 800 |
},
|
| 801 |
{
|
| 802 |
+
"epoch": 1.6,
|
| 803 |
+
"eval_loss": 0.2884255647659302,
|
| 804 |
+
"eval_na_accuracy": 0.94,
|
| 805 |
+
"eval_ordinal_accuracy": 0.5617070357554786,
|
| 806 |
+
"eval_runtime": 39.4898,
|
| 807 |
+
"eval_samples_per_second": 25.323,
|
| 808 |
+
"eval_steps_per_second": 3.165,
|
| 809 |
"step": 1000
|
| 810 |
},
|
| 811 |
{
|
| 812 |
+
"epoch": 1.62,
|
| 813 |
+
"grad_norm": 1.1306260824203491,
|
| 814 |
+
"learning_rate": 0.00011936000000000001,
|
| 815 |
+
"loss": 0.1595,
|
| 816 |
"step": 1010
|
| 817 |
},
|
| 818 |
{
|
| 819 |
+
"epoch": 1.63,
|
| 820 |
+
"grad_norm": 1.8953267335891724,
|
| 821 |
+
"learning_rate": 0.00011856,
|
| 822 |
+
"loss": 0.2489,
|
| 823 |
"step": 1020
|
| 824 |
},
|
| 825 |
{
|
| 826 |
+
"epoch": 1.65,
|
| 827 |
+
"grad_norm": 0.7074128985404968,
|
| 828 |
+
"learning_rate": 0.00011776,
|
| 829 |
+
"loss": 0.2485,
|
| 830 |
"step": 1030
|
| 831 |
},
|
| 832 |
{
|
| 833 |
+
"epoch": 1.66,
|
| 834 |
+
"grad_norm": 0.7052355408668518,
|
| 835 |
+
"learning_rate": 0.00011696,
|
| 836 |
+
"loss": 0.2075,
|
| 837 |
"step": 1040
|
| 838 |
},
|
| 839 |
{
|
| 840 |
+
"epoch": 1.68,
|
| 841 |
+
"grad_norm": 0.7830259203910828,
|
| 842 |
+
"learning_rate": 0.00011616,
|
| 843 |
+
"loss": 0.2346,
|
| 844 |
"step": 1050
|
| 845 |
},
|
| 846 |
{
|
| 847 |
+
"epoch": 1.7,
|
| 848 |
+
"grad_norm": 0.5882430672645569,
|
| 849 |
+
"learning_rate": 0.00011536000000000001,
|
| 850 |
+
"loss": 0.2136,
|
| 851 |
"step": 1060
|
| 852 |
},
|
| 853 |
{
|
| 854 |
+
"epoch": 1.71,
|
| 855 |
+
"grad_norm": 1.0235962867736816,
|
| 856 |
+
"learning_rate": 0.00011456,
|
| 857 |
+
"loss": 0.1753,
|
| 858 |
"step": 1070
|
| 859 |
},
|
| 860 |
{
|
| 861 |
+
"epoch": 1.73,
|
| 862 |
+
"grad_norm": 0.9401603937149048,
|
| 863 |
+
"learning_rate": 0.00011376,
|
| 864 |
+
"loss": 0.173,
|
| 865 |
"step": 1080
|
| 866 |
},
|
| 867 |
{
|
| 868 |
+
"epoch": 1.74,
|
| 869 |
+
"grad_norm": 1.0735399723052979,
|
| 870 |
+
"learning_rate": 0.00011296,
|
| 871 |
+
"loss": 0.1993,
|
| 872 |
"step": 1090
|
| 873 |
},
|
| 874 |
{
|
| 875 |
+
"epoch": 1.76,
|
| 876 |
+
"grad_norm": 0.6592912673950195,
|
| 877 |
+
"learning_rate": 0.00011216,
|
| 878 |
+
"loss": 0.1826,
|
| 879 |
"step": 1100
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 1.76,
|
| 883 |
+
"eval_loss": 0.2869604229927063,
|
| 884 |
+
"eval_na_accuracy": 0.943,
|
| 885 |
+
"eval_ordinal_accuracy": 0.6043829296424452,
|
| 886 |
+
"eval_runtime": 39.3069,
|
| 887 |
+
"eval_samples_per_second": 25.441,
|
| 888 |
+
"eval_steps_per_second": 3.18,
|
| 889 |
"step": 1100
|
| 890 |
},
|
| 891 |
{
|
| 892 |
+
"epoch": 1.78,
|
| 893 |
+
"grad_norm": 2.1649601459503174,
|
| 894 |
+
"learning_rate": 0.00011135999999999999,
|
| 895 |
+
"loss": 0.221,
|
| 896 |
"step": 1110
|
| 897 |
},
|
| 898 |
{
|
| 899 |
+
"epoch": 1.79,
|
| 900 |
+
"grad_norm": 1.7881801128387451,
|
| 901 |
+
"learning_rate": 0.00011056,
|
| 902 |
+
"loss": 0.1765,
|
| 903 |
"step": 1120
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 1.81,
|
| 907 |
+
"grad_norm": 1.3527191877365112,
|
| 908 |
+
"learning_rate": 0.00010975999999999999,
|
| 909 |
+
"loss": 0.1325,
|
| 910 |
"step": 1130
|
| 911 |
},
|
| 912 |
{
|
| 913 |
+
"epoch": 1.82,
|
| 914 |
+
"grad_norm": 0.7212499976158142,
|
| 915 |
+
"learning_rate": 0.00010896,
|
| 916 |
+
"loss": 0.2445,
|
| 917 |
"step": 1140
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 1.84,
|
| 921 |
+
"grad_norm": 0.9492518901824951,
|
| 922 |
+
"learning_rate": 0.00010816,
|
| 923 |
+
"loss": 0.2704,
|
| 924 |
"step": 1150
|
| 925 |
},
|
| 926 |
{
|
| 927 |
+
"epoch": 1.86,
|
| 928 |
+
"grad_norm": 0.4344118535518646,
|
| 929 |
+
"learning_rate": 0.00010736000000000002,
|
| 930 |
+
"loss": 0.1624,
|
| 931 |
"step": 1160
|
| 932 |
},
|
| 933 |
{
|
| 934 |
+
"epoch": 1.87,
|
| 935 |
+
"grad_norm": 0.4115823209285736,
|
| 936 |
+
"learning_rate": 0.00010656000000000001,
|
| 937 |
+
"loss": 0.2069,
|
| 938 |
"step": 1170
|
| 939 |
},
|
| 940 |
{
|
| 941 |
+
"epoch": 1.89,
|
| 942 |
+
"grad_norm": 0.6738015413284302,
|
| 943 |
+
"learning_rate": 0.00010576000000000002,
|
| 944 |
+
"loss": 0.208,
|
| 945 |
"step": 1180
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 1.9,
|
| 949 |
+
"grad_norm": 0.9090007543563843,
|
| 950 |
+
"learning_rate": 0.00010496000000000001,
|
| 951 |
+
"loss": 0.1793,
|
| 952 |
"step": 1190
|
| 953 |
},
|
| 954 |
{
|
| 955 |
+
"epoch": 1.92,
|
| 956 |
+
"grad_norm": 1.1480025053024292,
|
| 957 |
+
"learning_rate": 0.00010416000000000002,
|
| 958 |
+
"loss": 0.2013,
|
| 959 |
"step": 1200
|
| 960 |
},
|
| 961 |
{
|
| 962 |
+
"epoch": 1.92,
|
| 963 |
+
"eval_loss": 0.29365527629852295,
|
| 964 |
+
"eval_na_accuracy": 0.941,
|
| 965 |
+
"eval_ordinal_accuracy": 0.5905420991926182,
|
| 966 |
+
"eval_runtime": 39.6842,
|
| 967 |
+
"eval_samples_per_second": 25.199,
|
| 968 |
+
"eval_steps_per_second": 3.15,
|
| 969 |
"step": 1200
|
| 970 |
},
|
| 971 |
{
|
| 972 |
+
"epoch": 1.94,
|
| 973 |
+
"grad_norm": 0.62380051612854,
|
| 974 |
+
"learning_rate": 0.00010336000000000001,
|
| 975 |
+
"loss": 0.192,
|
| 976 |
"step": 1210
|
| 977 |
},
|
| 978 |
{
|
| 979 |
+
"epoch": 1.95,
|
| 980 |
+
"grad_norm": 0.9949710965156555,
|
| 981 |
+
"learning_rate": 0.00010256000000000001,
|
| 982 |
+
"loss": 0.226,
|
| 983 |
"step": 1220
|
| 984 |
},
|
| 985 |
{
|
| 986 |
+
"epoch": 1.97,
|
| 987 |
+
"grad_norm": 1.0634446144104004,
|
| 988 |
+
"learning_rate": 0.00010176000000000002,
|
| 989 |
+
"loss": 0.142,
|
| 990 |
"step": 1230
|
| 991 |
},
|
| 992 |
{
|
| 993 |
+
"epoch": 1.98,
|
| 994 |
+
"grad_norm": 0.8875225782394409,
|
| 995 |
+
"learning_rate": 0.00010096000000000001,
|
| 996 |
+
"loss": 0.1729,
|
| 997 |
"step": 1240
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
+
"epoch": 2.0,
|
| 1001 |
+
"grad_norm": 0.6193259358406067,
|
| 1002 |
+
"learning_rate": 0.00010016,
|
| 1003 |
+
"loss": 0.102,
|
| 1004 |
"step": 1250
|
| 1005 |
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.02,
|
| 1008 |
+
"grad_norm": 0.4042517840862274,
|
| 1009 |
+
"learning_rate": 9.936000000000001e-05,
|
| 1010 |
+
"loss": 0.0976,
|
| 1011 |
+
"step": 1260
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.03,
|
| 1015 |
+
"grad_norm": 0.4051195979118347,
|
| 1016 |
+
"learning_rate": 9.856e-05,
|
| 1017 |
+
"loss": 0.1448,
|
| 1018 |
+
"step": 1270
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.05,
|
| 1022 |
+
"grad_norm": 0.46061789989471436,
|
| 1023 |
+
"learning_rate": 9.776000000000001e-05,
|
| 1024 |
+
"loss": 0.0768,
|
| 1025 |
+
"step": 1280
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.06,
|
| 1029 |
+
"grad_norm": 0.5934004783630371,
|
| 1030 |
+
"learning_rate": 9.696000000000001e-05,
|
| 1031 |
+
"loss": 0.1404,
|
| 1032 |
+
"step": 1290
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.08,
|
| 1036 |
+
"grad_norm": 0.6819984316825867,
|
| 1037 |
+
"learning_rate": 9.616e-05,
|
| 1038 |
+
"loss": 0.0663,
|
| 1039 |
+
"step": 1300
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.08,
|
| 1043 |
+
"eval_loss": 0.2954486608505249,
|
| 1044 |
+
"eval_na_accuracy": 0.938,
|
| 1045 |
+
"eval_ordinal_accuracy": 0.6251441753171857,
|
| 1046 |
+
"eval_runtime": 39.658,
|
| 1047 |
+
"eval_samples_per_second": 25.216,
|
| 1048 |
+
"eval_steps_per_second": 3.152,
|
| 1049 |
+
"step": 1300
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.1,
|
| 1053 |
+
"grad_norm": 0.5849266052246094,
|
| 1054 |
+
"learning_rate": 9.536000000000001e-05,
|
| 1055 |
+
"loss": 0.1574,
|
| 1056 |
+
"step": 1310
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.11,
|
| 1060 |
+
"grad_norm": 0.9393780827522278,
|
| 1061 |
+
"learning_rate": 9.456e-05,
|
| 1062 |
+
"loss": 0.0979,
|
| 1063 |
+
"step": 1320
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.13,
|
| 1067 |
+
"grad_norm": 0.47529059648513794,
|
| 1068 |
+
"learning_rate": 9.376e-05,
|
| 1069 |
+
"loss": 0.1073,
|
| 1070 |
+
"step": 1330
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.14,
|
| 1074 |
+
"grad_norm": 0.4079722464084625,
|
| 1075 |
+
"learning_rate": 9.296e-05,
|
| 1076 |
+
"loss": 0.0868,
|
| 1077 |
+
"step": 1340
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.16,
|
| 1081 |
+
"grad_norm": 0.7292589545249939,
|
| 1082 |
+
"learning_rate": 9.216e-05,
|
| 1083 |
+
"loss": 0.1446,
|
| 1084 |
+
"step": 1350
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.18,
|
| 1088 |
+
"grad_norm": 0.9205511212348938,
|
| 1089 |
+
"learning_rate": 9.136e-05,
|
| 1090 |
+
"loss": 0.0907,
|
| 1091 |
+
"step": 1360
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.19,
|
| 1095 |
+
"grad_norm": 0.9218105673789978,
|
| 1096 |
+
"learning_rate": 9.056e-05,
|
| 1097 |
+
"loss": 0.1387,
|
| 1098 |
+
"step": 1370
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.21,
|
| 1102 |
+
"grad_norm": 0.5730422139167786,
|
| 1103 |
+
"learning_rate": 8.976e-05,
|
| 1104 |
+
"loss": 0.0882,
|
| 1105 |
+
"step": 1380
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.22,
|
| 1109 |
+
"grad_norm": 0.6922823190689087,
|
| 1110 |
+
"learning_rate": 8.896e-05,
|
| 1111 |
+
"loss": 0.0741,
|
| 1112 |
+
"step": 1390
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.24,
|
| 1116 |
+
"grad_norm": 1.1872971057891846,
|
| 1117 |
+
"learning_rate": 8.816000000000001e-05,
|
| 1118 |
+
"loss": 0.1503,
|
| 1119 |
+
"step": 1400
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.24,
|
| 1123 |
+
"eval_loss": 0.3187769651412964,
|
| 1124 |
+
"eval_na_accuracy": 0.937,
|
| 1125 |
+
"eval_ordinal_accuracy": 0.5986159169550173,
|
| 1126 |
+
"eval_runtime": 39.7243,
|
| 1127 |
+
"eval_samples_per_second": 25.173,
|
| 1128 |
+
"eval_steps_per_second": 3.147,
|
| 1129 |
+
"step": 1400
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.26,
|
| 1133 |
+
"grad_norm": 2.3350443840026855,
|
| 1134 |
+
"learning_rate": 8.736e-05,
|
| 1135 |
+
"loss": 0.1083,
|
| 1136 |
+
"step": 1410
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.27,
|
| 1140 |
+
"grad_norm": 0.8266046643257141,
|
| 1141 |
+
"learning_rate": 8.656000000000001e-05,
|
| 1142 |
+
"loss": 0.0684,
|
| 1143 |
+
"step": 1420
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.29,
|
| 1147 |
+
"grad_norm": 0.41480687260627747,
|
| 1148 |
+
"learning_rate": 8.576e-05,
|
| 1149 |
+
"loss": 0.0809,
|
| 1150 |
+
"step": 1430
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.3,
|
| 1154 |
+
"grad_norm": 0.4657377600669861,
|
| 1155 |
+
"learning_rate": 8.496e-05,
|
| 1156 |
+
"loss": 0.0718,
|
| 1157 |
+
"step": 1440
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.32,
|
| 1161 |
+
"grad_norm": 0.5419800877571106,
|
| 1162 |
+
"learning_rate": 8.416000000000001e-05,
|
| 1163 |
+
"loss": 0.1322,
|
| 1164 |
+
"step": 1450
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.34,
|
| 1168 |
+
"grad_norm": 1.167611837387085,
|
| 1169 |
+
"learning_rate": 8.336e-05,
|
| 1170 |
+
"loss": 0.1017,
|
| 1171 |
+
"step": 1460
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.35,
|
| 1175 |
+
"grad_norm": 0.449034184217453,
|
| 1176 |
+
"learning_rate": 8.256000000000001e-05,
|
| 1177 |
+
"loss": 0.0636,
|
| 1178 |
+
"step": 1470
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.37,
|
| 1182 |
+
"grad_norm": 0.6716451048851013,
|
| 1183 |
+
"learning_rate": 8.176e-05,
|
| 1184 |
+
"loss": 0.1109,
|
| 1185 |
+
"step": 1480
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.38,
|
| 1189 |
+
"grad_norm": 4.306596755981445,
|
| 1190 |
+
"learning_rate": 8.096e-05,
|
| 1191 |
+
"loss": 0.0898,
|
| 1192 |
+
"step": 1490
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.4,
|
| 1196 |
+
"grad_norm": 0.41288742423057556,
|
| 1197 |
+
"learning_rate": 8.016e-05,
|
| 1198 |
+
"loss": 0.0611,
|
| 1199 |
+
"step": 1500
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.4,
|
| 1203 |
+
"eval_loss": 0.33932703733444214,
|
| 1204 |
+
"eval_na_accuracy": 0.945,
|
| 1205 |
+
"eval_ordinal_accuracy": 0.5997693194925029,
|
| 1206 |
+
"eval_runtime": 39.236,
|
| 1207 |
+
"eval_samples_per_second": 25.487,
|
| 1208 |
+
"eval_steps_per_second": 3.186,
|
| 1209 |
+
"step": 1500
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 2.42,
|
| 1213 |
+
"grad_norm": 0.7951626181602478,
|
| 1214 |
+
"learning_rate": 7.936e-05,
|
| 1215 |
+
"loss": 0.0799,
|
| 1216 |
+
"step": 1510
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 2.43,
|
| 1220 |
+
"grad_norm": 1.0197049379348755,
|
| 1221 |
+
"learning_rate": 7.856000000000001e-05,
|
| 1222 |
+
"loss": 0.0928,
|
| 1223 |
+
"step": 1520
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 2.45,
|
| 1227 |
+
"grad_norm": 0.6486759185791016,
|
| 1228 |
+
"learning_rate": 7.776e-05,
|
| 1229 |
+
"loss": 0.0964,
|
| 1230 |
+
"step": 1530
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 2.46,
|
| 1234 |
+
"grad_norm": 1.0220657587051392,
|
| 1235 |
+
"learning_rate": 7.696e-05,
|
| 1236 |
+
"loss": 0.1736,
|
| 1237 |
+
"step": 1540
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 2.48,
|
| 1241 |
+
"grad_norm": 2.3006441593170166,
|
| 1242 |
+
"learning_rate": 7.616e-05,
|
| 1243 |
+
"loss": 0.0993,
|
| 1244 |
+
"step": 1550
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.5,
|
| 1248 |
+
"grad_norm": 0.4701670706272125,
|
| 1249 |
+
"learning_rate": 7.536000000000001e-05,
|
| 1250 |
+
"loss": 0.0779,
|
| 1251 |
+
"step": 1560
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 2.51,
|
| 1255 |
+
"grad_norm": 0.641832172870636,
|
| 1256 |
+
"learning_rate": 7.456e-05,
|
| 1257 |
+
"loss": 0.0586,
|
| 1258 |
+
"step": 1570
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 2.53,
|
| 1262 |
+
"grad_norm": 0.5836305618286133,
|
| 1263 |
+
"learning_rate": 7.376000000000001e-05,
|
| 1264 |
+
"loss": 0.053,
|
| 1265 |
+
"step": 1580
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 2.54,
|
| 1269 |
+
"grad_norm": 0.6500815153121948,
|
| 1270 |
+
"learning_rate": 7.296e-05,
|
| 1271 |
+
"loss": 0.0779,
|
| 1272 |
+
"step": 1590
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 2.56,
|
| 1276 |
+
"grad_norm": 0.5682386755943298,
|
| 1277 |
+
"learning_rate": 7.216e-05,
|
| 1278 |
+
"loss": 0.0743,
|
| 1279 |
+
"step": 1600
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 2.56,
|
| 1283 |
+
"eval_loss": 0.3182476758956909,
|
| 1284 |
+
"eval_na_accuracy": 0.942,
|
| 1285 |
+
"eval_ordinal_accuracy": 0.6482122260668973,
|
| 1286 |
+
"eval_runtime": 40.5507,
|
| 1287 |
+
"eval_samples_per_second": 24.66,
|
| 1288 |
+
"eval_steps_per_second": 3.083,
|
| 1289 |
+
"step": 1600
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 2.58,
|
| 1293 |
+
"grad_norm": 0.5527540445327759,
|
| 1294 |
+
"learning_rate": 7.136000000000001e-05,
|
| 1295 |
+
"loss": 0.094,
|
| 1296 |
+
"step": 1610
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 2.59,
|
| 1300 |
+
"grad_norm": 0.8710426092147827,
|
| 1301 |
+
"learning_rate": 7.056e-05,
|
| 1302 |
+
"loss": 0.0839,
|
| 1303 |
+
"step": 1620
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 2.61,
|
| 1307 |
+
"grad_norm": 0.9312260746955872,
|
| 1308 |
+
"learning_rate": 6.976000000000001e-05,
|
| 1309 |
+
"loss": 0.155,
|
| 1310 |
+
"step": 1630
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.62,
|
| 1314 |
+
"grad_norm": 0.48695412278175354,
|
| 1315 |
+
"learning_rate": 6.896e-05,
|
| 1316 |
+
"loss": 0.0606,
|
| 1317 |
+
"step": 1640
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 2.64,
|
| 1321 |
+
"grad_norm": 0.525652289390564,
|
| 1322 |
+
"learning_rate": 6.816e-05,
|
| 1323 |
+
"loss": 0.0715,
|
| 1324 |
+
"step": 1650
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 2.66,
|
| 1328 |
+
"grad_norm": 0.7670960426330566,
|
| 1329 |
+
"learning_rate": 6.736e-05,
|
| 1330 |
+
"loss": 0.1108,
|
| 1331 |
+
"step": 1660
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 2.67,
|
| 1335 |
+
"grad_norm": 1.0041375160217285,
|
| 1336 |
+
"learning_rate": 6.656e-05,
|
| 1337 |
+
"loss": 0.1257,
|
| 1338 |
+
"step": 1670
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 2.69,
|
| 1342 |
+
"grad_norm": 0.3819805681705475,
|
| 1343 |
+
"learning_rate": 6.576e-05,
|
| 1344 |
+
"loss": 0.1004,
|
| 1345 |
+
"step": 1680
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 2.7,
|
| 1349 |
+
"grad_norm": 0.5372006893157959,
|
| 1350 |
+
"learning_rate": 6.496e-05,
|
| 1351 |
+
"loss": 0.0825,
|
| 1352 |
+
"step": 1690
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 2.72,
|
| 1356 |
+
"grad_norm": 0.5835949182510376,
|
| 1357 |
+
"learning_rate": 6.416e-05,
|
| 1358 |
+
"loss": 0.0908,
|
| 1359 |
+
"step": 1700
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 2.72,
|
| 1363 |
+
"eval_loss": 0.3332485854625702,
|
| 1364 |
+
"eval_na_accuracy": 0.942,
|
| 1365 |
+
"eval_ordinal_accuracy": 0.6482122260668973,
|
| 1366 |
+
"eval_runtime": 39.9642,
|
| 1367 |
+
"eval_samples_per_second": 25.022,
|
| 1368 |
+
"eval_steps_per_second": 3.128,
|
| 1369 |
+
"step": 1700
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 2.74,
|
| 1373 |
+
"grad_norm": 0.677947461605072,
|
| 1374 |
+
"learning_rate": 6.336e-05,
|
| 1375 |
+
"loss": 0.1086,
|
| 1376 |
+
"step": 1710
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 2.75,
|
| 1380 |
+
"grad_norm": 0.7373325228691101,
|
| 1381 |
+
"learning_rate": 6.256000000000001e-05,
|
| 1382 |
+
"loss": 0.0698,
|
| 1383 |
+
"step": 1720
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 2.77,
|
| 1387 |
+
"grad_norm": 0.7738047242164612,
|
| 1388 |
+
"learning_rate": 6.176e-05,
|
| 1389 |
+
"loss": 0.1118,
|
| 1390 |
+
"step": 1730
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 2.78,
|
| 1394 |
+
"grad_norm": 2.052891254425049,
|
| 1395 |
+
"learning_rate": 6.0960000000000006e-05,
|
| 1396 |
+
"loss": 0.1002,
|
| 1397 |
+
"step": 1740
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 2.8,
|
| 1401 |
+
"grad_norm": 0.26311352849006653,
|
| 1402 |
+
"learning_rate": 6.016000000000001e-05,
|
| 1403 |
+
"loss": 0.0944,
|
| 1404 |
+
"step": 1750
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 2.82,
|
| 1408 |
+
"grad_norm": 0.8190409541130066,
|
| 1409 |
+
"learning_rate": 5.936000000000001e-05,
|
| 1410 |
+
"loss": 0.0717,
|
| 1411 |
+
"step": 1760
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 2.83,
|
| 1415 |
+
"grad_norm": 0.5824436545372009,
|
| 1416 |
+
"learning_rate": 5.856e-05,
|
| 1417 |
+
"loss": 0.0746,
|
| 1418 |
+
"step": 1770
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 2.85,
|
| 1422 |
+
"grad_norm": 0.5489352941513062,
|
| 1423 |
+
"learning_rate": 5.776e-05,
|
| 1424 |
+
"loss": 0.1063,
|
| 1425 |
+
"step": 1780
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 2.86,
|
| 1429 |
+
"grad_norm": 0.656225323677063,
|
| 1430 |
+
"learning_rate": 5.6960000000000004e-05,
|
| 1431 |
+
"loss": 0.0763,
|
| 1432 |
+
"step": 1790
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.88,
|
| 1436 |
+
"grad_norm": 0.8495000600814819,
|
| 1437 |
+
"learning_rate": 5.6160000000000004e-05,
|
| 1438 |
+
"loss": 0.1108,
|
| 1439 |
+
"step": 1800
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 2.88,
|
| 1443 |
+
"eval_loss": 0.32561835646629333,
|
| 1444 |
+
"eval_na_accuracy": 0.943,
|
| 1445 |
+
"eval_ordinal_accuracy": 0.6459054209919262,
|
| 1446 |
+
"eval_runtime": 39.4673,
|
| 1447 |
+
"eval_samples_per_second": 25.337,
|
| 1448 |
+
"eval_steps_per_second": 3.167,
|
| 1449 |
+
"step": 1800
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 2.9,
|
| 1453 |
+
"grad_norm": 0.617258608341217,
|
| 1454 |
+
"learning_rate": 5.536e-05,
|
| 1455 |
+
"loss": 0.1203,
|
| 1456 |
+
"step": 1810
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.91,
|
| 1460 |
+
"grad_norm": 0.4484919011592865,
|
| 1461 |
+
"learning_rate": 5.456e-05,
|
| 1462 |
+
"loss": 0.0573,
|
| 1463 |
+
"step": 1820
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 2.93,
|
| 1467 |
+
"grad_norm": 0.533388614654541,
|
| 1468 |
+
"learning_rate": 5.376e-05,
|
| 1469 |
+
"loss": 0.0762,
|
| 1470 |
+
"step": 1830
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.94,
|
| 1474 |
+
"grad_norm": 0.4078121483325958,
|
| 1475 |
+
"learning_rate": 5.296e-05,
|
| 1476 |
+
"loss": 0.0643,
|
| 1477 |
+
"step": 1840
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 2.96,
|
| 1481 |
+
"grad_norm": 0.5678732395172119,
|
| 1482 |
+
"learning_rate": 5.2159999999999995e-05,
|
| 1483 |
+
"loss": 0.1126,
|
| 1484 |
+
"step": 1850
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 2.98,
|
| 1488 |
+
"grad_norm": 0.6543716192245483,
|
| 1489 |
+
"learning_rate": 5.1359999999999996e-05,
|
| 1490 |
+
"loss": 0.0763,
|
| 1491 |
+
"step": 1860
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 2.99,
|
| 1495 |
+
"grad_norm": 0.6005885601043701,
|
| 1496 |
+
"learning_rate": 5.056000000000001e-05,
|
| 1497 |
+
"loss": 0.0949,
|
| 1498 |
+
"step": 1870
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 3.01,
|
| 1502 |
+
"grad_norm": 0.4467845559120178,
|
| 1503 |
+
"learning_rate": 4.976e-05,
|
| 1504 |
+
"loss": 0.0481,
|
| 1505 |
+
"step": 1880
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 3.02,
|
| 1509 |
+
"grad_norm": 0.48746606707572937,
|
| 1510 |
+
"learning_rate": 4.896e-05,
|
| 1511 |
+
"loss": 0.0415,
|
| 1512 |
+
"step": 1890
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 3.04,
|
| 1516 |
+
"grad_norm": 0.7011713981628418,
|
| 1517 |
+
"learning_rate": 4.816e-05,
|
| 1518 |
+
"loss": 0.0786,
|
| 1519 |
+
"step": 1900
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 3.04,
|
| 1523 |
+
"eval_loss": 0.3222349286079407,
|
| 1524 |
+
"eval_na_accuracy": 0.944,
|
| 1525 |
+
"eval_ordinal_accuracy": 0.6539792387543253,
|
| 1526 |
+
"eval_runtime": 39.2271,
|
| 1527 |
+
"eval_samples_per_second": 25.493,
|
| 1528 |
+
"eval_steps_per_second": 3.187,
|
| 1529 |
+
"step": 1900
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 3.06,
|
| 1533 |
+
"grad_norm": 0.5153183341026306,
|
| 1534 |
+
"learning_rate": 4.736000000000001e-05,
|
| 1535 |
+
"loss": 0.0387,
|
| 1536 |
+
"step": 1910
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 3.07,
|
| 1540 |
+
"grad_norm": 0.43751129508018494,
|
| 1541 |
+
"learning_rate": 4.656e-05,
|
| 1542 |
+
"loss": 0.0364,
|
| 1543 |
+
"step": 1920
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 3.09,
|
| 1547 |
+
"grad_norm": 0.7584701776504517,
|
| 1548 |
+
"learning_rate": 4.576e-05,
|
| 1549 |
+
"loss": 0.0397,
|
| 1550 |
+
"step": 1930
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 3.1,
|
| 1554 |
+
"grad_norm": 0.20170661807060242,
|
| 1555 |
+
"learning_rate": 4.496e-05,
|
| 1556 |
+
"loss": 0.0288,
|
| 1557 |
+
"step": 1940
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 3.12,
|
| 1561 |
+
"grad_norm": 0.2583639919757843,
|
| 1562 |
+
"learning_rate": 4.4160000000000004e-05,
|
| 1563 |
+
"loss": 0.0302,
|
| 1564 |
+
"step": 1950
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 3.14,
|
| 1568 |
+
"grad_norm": 3.9720704555511475,
|
| 1569 |
+
"learning_rate": 4.336e-05,
|
| 1570 |
+
"loss": 0.0484,
|
| 1571 |
+
"step": 1960
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.15,
|
| 1575 |
+
"grad_norm": 0.3367606997489929,
|
| 1576 |
+
"learning_rate": 4.256e-05,
|
| 1577 |
+
"loss": 0.0387,
|
| 1578 |
+
"step": 1970
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 3.17,
|
| 1582 |
+
"grad_norm": 0.7610962986946106,
|
| 1583 |
+
"learning_rate": 4.176000000000001e-05,
|
| 1584 |
+
"loss": 0.0424,
|
| 1585 |
+
"step": 1980
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 3.18,
|
| 1589 |
+
"grad_norm": 0.6901140213012695,
|
| 1590 |
+
"learning_rate": 4.096e-05,
|
| 1591 |
+
"loss": 0.035,
|
| 1592 |
+
"step": 1990
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 3.2,
|
| 1596 |
+
"grad_norm": 0.2873363196849823,
|
| 1597 |
+
"learning_rate": 4.016e-05,
|
| 1598 |
+
"loss": 0.043,
|
| 1599 |
+
"step": 2000
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 3.2,
|
| 1603 |
+
"eval_loss": 0.35012441873550415,
|
| 1604 |
+
"eval_na_accuracy": 0.941,
|
| 1605 |
+
"eval_ordinal_accuracy": 0.6482122260668973,
|
| 1606 |
+
"eval_runtime": 39.427,
|
| 1607 |
+
"eval_samples_per_second": 25.363,
|
| 1608 |
+
"eval_steps_per_second": 3.17,
|
| 1609 |
+
"step": 2000
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 3.22,
|
| 1613 |
+
"grad_norm": 0.5277103781700134,
|
| 1614 |
+
"learning_rate": 3.936e-05,
|
| 1615 |
+
"loss": 0.0423,
|
| 1616 |
+
"step": 2010
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 3.23,
|
| 1620 |
+
"grad_norm": 0.3088182508945465,
|
| 1621 |
+
"learning_rate": 3.8560000000000004e-05,
|
| 1622 |
+
"loss": 0.0305,
|
| 1623 |
+
"step": 2020
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 3.25,
|
| 1627 |
+
"grad_norm": 0.3621159791946411,
|
| 1628 |
+
"learning_rate": 3.776e-05,
|
| 1629 |
+
"loss": 0.0398,
|
| 1630 |
+
"step": 2030
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 3.26,
|
| 1634 |
+
"grad_norm": 0.6761226654052734,
|
| 1635 |
+
"learning_rate": 3.696e-05,
|
| 1636 |
+
"loss": 0.0606,
|
| 1637 |
+
"step": 2040
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 3.28,
|
| 1641 |
+
"grad_norm": 0.9860779047012329,
|
| 1642 |
+
"learning_rate": 3.616e-05,
|
| 1643 |
+
"loss": 0.0437,
|
| 1644 |
+
"step": 2050
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 3.3,
|
| 1648 |
+
"grad_norm": 0.6743874549865723,
|
| 1649 |
+
"learning_rate": 3.536000000000001e-05,
|
| 1650 |
+
"loss": 0.0328,
|
| 1651 |
+
"step": 2060
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 3.31,
|
| 1655 |
+
"grad_norm": 1.5928354263305664,
|
| 1656 |
+
"learning_rate": 3.456e-05,
|
| 1657 |
+
"loss": 0.0261,
|
| 1658 |
+
"step": 2070
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 3.33,
|
| 1662 |
+
"grad_norm": 0.9067389965057373,
|
| 1663 |
+
"learning_rate": 3.376e-05,
|
| 1664 |
+
"loss": 0.0268,
|
| 1665 |
+
"step": 2080
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 3.34,
|
| 1669 |
+
"grad_norm": 0.5733221173286438,
|
| 1670 |
+
"learning_rate": 3.296e-05,
|
| 1671 |
+
"loss": 0.0335,
|
| 1672 |
+
"step": 2090
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 3.36,
|
| 1676 |
+
"grad_norm": 0.7042862772941589,
|
| 1677 |
+
"learning_rate": 3.2160000000000004e-05,
|
| 1678 |
+
"loss": 0.0472,
|
| 1679 |
+
"step": 2100
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 3.36,
|
| 1683 |
+
"eval_loss": 0.34554365277290344,
|
| 1684 |
+
"eval_na_accuracy": 0.943,
|
| 1685 |
+
"eval_ordinal_accuracy": 0.6608996539792388,
|
| 1686 |
+
"eval_runtime": 40.1655,
|
| 1687 |
+
"eval_samples_per_second": 24.897,
|
| 1688 |
+
"eval_steps_per_second": 3.112,
|
| 1689 |
+
"step": 2100
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.38,
|
| 1693 |
+
"grad_norm": 0.8036783933639526,
|
| 1694 |
+
"learning_rate": 3.136e-05,
|
| 1695 |
+
"loss": 0.0402,
|
| 1696 |
+
"step": 2110
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 3.39,
|
| 1700 |
+
"grad_norm": 0.4863825738430023,
|
| 1701 |
+
"learning_rate": 3.056e-05,
|
| 1702 |
+
"loss": 0.0507,
|
| 1703 |
+
"step": 2120
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 3.41,
|
| 1707 |
+
"grad_norm": 0.5171158313751221,
|
| 1708 |
+
"learning_rate": 2.976e-05,
|
| 1709 |
+
"loss": 0.0333,
|
| 1710 |
+
"step": 2130
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 3.42,
|
| 1714 |
+
"grad_norm": 0.21965381503105164,
|
| 1715 |
+
"learning_rate": 2.8960000000000004e-05,
|
| 1716 |
+
"loss": 0.0277,
|
| 1717 |
+
"step": 2140
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.44,
|
| 1721 |
+
"grad_norm": 0.20841450989246368,
|
| 1722 |
+
"learning_rate": 2.816e-05,
|
| 1723 |
+
"loss": 0.0259,
|
| 1724 |
+
"step": 2150
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.46,
|
| 1728 |
+
"grad_norm": 0.5015869736671448,
|
| 1729 |
+
"learning_rate": 2.7360000000000002e-05,
|
| 1730 |
+
"loss": 0.0316,
|
| 1731 |
+
"step": 2160
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 3.47,
|
| 1735 |
+
"grad_norm": 0.7938678860664368,
|
| 1736 |
+
"learning_rate": 2.6560000000000003e-05,
|
| 1737 |
+
"loss": 0.0301,
|
| 1738 |
+
"step": 2170
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 3.49,
|
| 1742 |
+
"grad_norm": 0.44840845465660095,
|
| 1743 |
+
"learning_rate": 2.576e-05,
|
| 1744 |
+
"loss": 0.0406,
|
| 1745 |
+
"step": 2180
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 3.5,
|
| 1749 |
+
"grad_norm": 0.35510167479515076,
|
| 1750 |
+
"learning_rate": 2.496e-05,
|
| 1751 |
+
"loss": 0.0268,
|
| 1752 |
+
"step": 2190
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 3.52,
|
| 1756 |
+
"grad_norm": 0.37328681349754333,
|
| 1757 |
+
"learning_rate": 2.4160000000000002e-05,
|
| 1758 |
+
"loss": 0.032,
|
| 1759 |
+
"step": 2200
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 3.52,
|
| 1763 |
+
"eval_loss": 0.35616353154182434,
|
| 1764 |
+
"eval_na_accuracy": 0.94,
|
| 1765 |
+
"eval_ordinal_accuracy": 0.6516724336793541,
|
| 1766 |
+
"eval_runtime": 40.3825,
|
| 1767 |
+
"eval_samples_per_second": 24.763,
|
| 1768 |
+
"eval_steps_per_second": 3.095,
|
| 1769 |
+
"step": 2200
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 3.54,
|
| 1773 |
+
"grad_norm": 0.24070143699645996,
|
| 1774 |
+
"learning_rate": 2.336e-05,
|
| 1775 |
+
"loss": 0.0262,
|
| 1776 |
+
"step": 2210
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 3.55,
|
| 1780 |
+
"grad_norm": 1.0428861379623413,
|
| 1781 |
+
"learning_rate": 2.256e-05,
|
| 1782 |
+
"loss": 0.04,
|
| 1783 |
+
"step": 2220
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 3.57,
|
| 1787 |
+
"grad_norm": 0.626348614692688,
|
| 1788 |
+
"learning_rate": 2.176e-05,
|
| 1789 |
+
"loss": 0.0275,
|
| 1790 |
+
"step": 2230
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 3.58,
|
| 1794 |
+
"grad_norm": 0.47826460003852844,
|
| 1795 |
+
"learning_rate": 2.0960000000000003e-05,
|
| 1796 |
+
"loss": 0.0379,
|
| 1797 |
+
"step": 2240
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 3.6,
|
| 1801 |
+
"grad_norm": 2.685340166091919,
|
| 1802 |
+
"learning_rate": 2.016e-05,
|
| 1803 |
+
"loss": 0.047,
|
| 1804 |
+
"step": 2250
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 3.62,
|
| 1808 |
+
"grad_norm": 0.1495877057313919,
|
| 1809 |
+
"learning_rate": 1.936e-05,
|
| 1810 |
+
"loss": 0.0389,
|
| 1811 |
+
"step": 2260
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 3.63,
|
| 1815 |
+
"grad_norm": 0.5789759755134583,
|
| 1816 |
+
"learning_rate": 1.856e-05,
|
| 1817 |
+
"loss": 0.0337,
|
| 1818 |
+
"step": 2270
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 3.65,
|
| 1822 |
+
"grad_norm": 0.4255303144454956,
|
| 1823 |
+
"learning_rate": 1.7760000000000003e-05,
|
| 1824 |
+
"loss": 0.0282,
|
| 1825 |
+
"step": 2280
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 3.66,
|
| 1829 |
+
"grad_norm": 0.7483348846435547,
|
| 1830 |
+
"learning_rate": 1.696e-05,
|
| 1831 |
+
"loss": 0.0668,
|
| 1832 |
+
"step": 2290
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 3.68,
|
| 1836 |
+
"grad_norm": 0.23885439336299896,
|
| 1837 |
+
"learning_rate": 1.616e-05,
|
| 1838 |
+
"loss": 0.0434,
|
| 1839 |
+
"step": 2300
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 3.68,
|
| 1843 |
+
"eval_loss": 0.34990155696868896,
|
| 1844 |
+
"eval_na_accuracy": 0.94,
|
| 1845 |
+
"eval_ordinal_accuracy": 0.6597462514417531,
|
| 1846 |
+
"eval_runtime": 40.407,
|
| 1847 |
+
"eval_samples_per_second": 24.748,
|
| 1848 |
+
"eval_steps_per_second": 3.094,
|
| 1849 |
+
"step": 2300
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 3.7,
|
| 1853 |
+
"grad_norm": 0.3341818153858185,
|
| 1854 |
+
"learning_rate": 1.536e-05,
|
| 1855 |
+
"loss": 0.0358,
|
| 1856 |
+
"step": 2310
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 3.71,
|
| 1860 |
+
"grad_norm": 0.6008884310722351,
|
| 1861 |
+
"learning_rate": 1.4560000000000001e-05,
|
| 1862 |
+
"loss": 0.0394,
|
| 1863 |
+
"step": 2320
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 3.73,
|
| 1867 |
+
"grad_norm": 0.3966546654701233,
|
| 1868 |
+
"learning_rate": 1.376e-05,
|
| 1869 |
+
"loss": 0.0346,
|
| 1870 |
+
"step": 2330
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 3.74,
|
| 1874 |
+
"grad_norm": 0.46933791041374207,
|
| 1875 |
+
"learning_rate": 1.296e-05,
|
| 1876 |
+
"loss": 0.0227,
|
| 1877 |
+
"step": 2340
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 3.76,
|
| 1881 |
+
"grad_norm": 0.6652282476425171,
|
| 1882 |
+
"learning_rate": 1.216e-05,
|
| 1883 |
+
"loss": 0.0393,
|
| 1884 |
+
"step": 2350
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 3.78,
|
| 1888 |
+
"grad_norm": 0.23938482999801636,
|
| 1889 |
+
"learning_rate": 1.1360000000000001e-05,
|
| 1890 |
+
"loss": 0.0267,
|
| 1891 |
+
"step": 2360
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 3.79,
|
| 1895 |
+
"grad_norm": 0.6050881147384644,
|
| 1896 |
+
"learning_rate": 1.056e-05,
|
| 1897 |
+
"loss": 0.0287,
|
| 1898 |
+
"step": 2370
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 3.81,
|
| 1902 |
+
"grad_norm": 0.22671189904212952,
|
| 1903 |
+
"learning_rate": 9.760000000000001e-06,
|
| 1904 |
+
"loss": 0.0491,
|
| 1905 |
+
"step": 2380
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 3.82,
|
| 1909 |
+
"grad_norm": 0.5296955704689026,
|
| 1910 |
+
"learning_rate": 8.96e-06,
|
| 1911 |
+
"loss": 0.0266,
|
| 1912 |
+
"step": 2390
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 3.84,
|
| 1916 |
+
"grad_norm": 0.5424560308456421,
|
| 1917 |
+
"learning_rate": 8.160000000000001e-06,
|
| 1918 |
+
"loss": 0.0341,
|
| 1919 |
+
"step": 2400
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 3.84,
|
| 1923 |
+
"eval_loss": 0.3610887825489044,
|
| 1924 |
+
"eval_na_accuracy": 0.94,
|
| 1925 |
+
"eval_ordinal_accuracy": 0.6482122260668973,
|
| 1926 |
+
"eval_runtime": 40.741,
|
| 1927 |
+
"eval_samples_per_second": 24.545,
|
| 1928 |
+
"eval_steps_per_second": 3.068,
|
| 1929 |
+
"step": 2400
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 3.86,
|
| 1933 |
+
"grad_norm": 0.4790421724319458,
|
| 1934 |
+
"learning_rate": 7.36e-06,
|
| 1935 |
+
"loss": 0.0319,
|
| 1936 |
+
"step": 2410
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 3.87,
|
| 1940 |
+
"grad_norm": 0.4021483063697815,
|
| 1941 |
+
"learning_rate": 6.560000000000001e-06,
|
| 1942 |
+
"loss": 0.0551,
|
| 1943 |
+
"step": 2420
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 3.89,
|
| 1947 |
+
"grad_norm": 0.43051794171333313,
|
| 1948 |
+
"learning_rate": 5.76e-06,
|
| 1949 |
+
"loss": 0.0281,
|
| 1950 |
+
"step": 2430
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 3.9,
|
| 1954 |
+
"grad_norm": 0.23781944811344147,
|
| 1955 |
+
"learning_rate": 4.96e-06,
|
| 1956 |
+
"loss": 0.0306,
|
| 1957 |
+
"step": 2440
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 3.92,
|
| 1961 |
+
"grad_norm": 0.6060004234313965,
|
| 1962 |
+
"learning_rate": 4.16e-06,
|
| 1963 |
+
"loss": 0.0326,
|
| 1964 |
+
"step": 2450
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 3.94,
|
| 1968 |
+
"grad_norm": 0.5149852633476257,
|
| 1969 |
+
"learning_rate": 3.36e-06,
|
| 1970 |
+
"loss": 0.0266,
|
| 1971 |
+
"step": 2460
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 3.95,
|
| 1975 |
+
"grad_norm": 0.579931914806366,
|
| 1976 |
+
"learning_rate": 2.56e-06,
|
| 1977 |
+
"loss": 0.0236,
|
| 1978 |
+
"step": 2470
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 3.97,
|
| 1982 |
+
"grad_norm": 0.14379101991653442,
|
| 1983 |
+
"learning_rate": 1.76e-06,
|
| 1984 |
+
"loss": 0.0221,
|
| 1985 |
+
"step": 2480
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 3.98,
|
| 1989 |
+
"grad_norm": 0.5184658765792847,
|
| 1990 |
+
"learning_rate": 9.6e-07,
|
| 1991 |
+
"loss": 0.0281,
|
| 1992 |
+
"step": 2490
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 4.0,
|
| 1996 |
+
"grad_norm": 0.5299363732337952,
|
| 1997 |
+
"learning_rate": 1.6e-07,
|
| 1998 |
+
"loss": 0.0305,
|
| 1999 |
+
"step": 2500
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 4.0,
|
| 2003 |
+
"eval_loss": 0.36354970932006836,
|
| 2004 |
+
"eval_na_accuracy": 0.939,
|
| 2005 |
+
"eval_ordinal_accuracy": 0.6608996539792388,
|
| 2006 |
+
"eval_runtime": 40.233,
|
| 2007 |
+
"eval_samples_per_second": 24.855,
|
| 2008 |
+
"eval_steps_per_second": 3.107,
|
| 2009 |
+
"step": 2500
|
| 2010 |
+
},
|
| 2011 |
{
|
| 2012 |
"epoch": 4.0,
|
| 2013 |
+
"step": 2500,
|
| 2014 |
+
"total_flos": 3.0997907103744e+18,
|
| 2015 |
+
"train_loss": 0.15650403581261635,
|
| 2016 |
+
"train_runtime": 3981.9696,
|
| 2017 |
+
"train_samples_per_second": 10.045,
|
| 2018 |
+
"train_steps_per_second": 0.628
|
| 2019 |
}
|
| 2020 |
],
|
| 2021 |
"logging_steps": 10,
|
| 2022 |
+
"max_steps": 2500,
|
| 2023 |
"num_input_tokens_seen": 0,
|
| 2024 |
"num_train_epochs": 4,
|
| 2025 |
"save_steps": 100,
|
| 2026 |
+
"total_flos": 3.0997907103744e+18,
|
| 2027 |
"train_batch_size": 16,
|
| 2028 |
"trial_name": null,
|
| 2029 |
"trial_params": null
|